Evaluation of the Secondary Triage Pilots in the Ambulance Communication Centres Final Report

Transcription

1 Evaluation of the Secondary Triage Pilots in the Ambulance Communication Centres Final Report Prepared for: Prepared by: Contract held by: Ministry of Health Rae Torrie, Robyn Bailey, Julian King, Gary Strong with Jeni Irving and Sandy Dawson Evaluation Works Ltd

2 Table of contents Executive summary... 1 Introduction... 6 Background... 6 The secondary triage pilots... 6 Strategic overview... 7 Delivery processes... 7 Evaluation brief... 8 Methodology... 9 Key evaluation questions (KEQs)...9 Evaluation design...9 Methods Stakeholder interviews Telephone survey to service users Clinical safety review of calls Desktop data analysis Analysis Findings KEQ 1: How well are the secondary triage processes working in each pilot site? Overall finding (level of effectiveness against rubric 1) Triage tools Range of dispositions Recruitment, selection and training of personnel Clinical governance and risk management Clinical safety Seamless interface Timeliness Credibility and acceptability to service users Credibility and acceptability to health professionals Mechanisms for learning and improving KEQ 2: How well are the secondary triage pilots achieving their intended outcomes? Overall finding (level of effectiveness against rubric 2) Clinical safety Level of care to which people are triaged Impact on ambulance dispatch, Impact on emergency departments KEQ 3: What are the efficiency implications of the pilot processes and outcomes? Operational costs of call triaging Impacts on wider health system costs Implications for health system efficiency KEQ 4: What can be learned from the evaluation to inform future decisions? Success factors Considerations for the future of secondary triage Appendices Appendix One: Process maps of secondary triage in practice Appendix Two: KEQs, rubrics and exploratory questions Appendix Three: Phone survey of users of secondary triage i

3 Acknowledgements: The evaluation team gratefully acknowledges the willing and positive assistance of all three secondary triage pilot providers St John, Wellington Free Ambulance and Medibank Health Solutions - in facilitating arrangements for visits, providing information and generally supporting the evaluation process. Disclaimer: The information in this report is presented in good faith using the information available to us at the time of preparation. It is provided on the basis that the authors of the report are not liable to any person or organisation for any damage or loss which may occur in relation to taking or not taking action in respect of any information or advice within this proposal. Citation: Any content reproduced from this report should be cited as follows: Torrie, R., Bailey, R., King, J., Strong, G., Irving, J., and Dawson, A. (2013). Evaluation of the Secondary Triage Pilots in the Ambulance Communications Centres. Prepared for the Ministry of Health by Evaluation Works Limited. ii

4 Executive summary The Ministry of Health (MoH) contracted Evaluation Works Ltd to conduct an evaluation of two secondary triage pilots. The evaluation covered a five-month period, from the start of the pilots in Oct 2012 till end Feb The objectives of the pilots are to: 1. Provide people calling 111 for an ambulance with the lowest safe level of care appropriate to their clinical need 2. Contribute to the cost-effectiveness of that care, for example by improved utilization of current ambulance resources. The secondary triage pilots are part of the Urgent and unplanned care integration project. There is interest in the extent to which the pilots, by changing service delivery arrangements, contribute to achieving more integrated care. The secondary triage pilots were undertaken by St John in the St John Canterbury dispatch area, and Wellington Free Ambulance (WFA) and Medibank Health Solutions NZ Ltd in WFA s dispatch area which covers the greater Wellington region and Wairarapa. The pilots focus on 111 calls that are deemed low acuity and appropriate for secondary triage. Transfer of 111 calls to secondary triage means that instead of an emergency ambulance being immediately dispatched, further clinical triage is provided for callers, and appropriate advice given on the health service they should access and in what timeframe. This includes the option of an ambulance if needed. The evaluation addressed four key questions: 1. How well are the secondary triage processes working in each pilot site? 2. How well are the secondary triage pilots achieving their intended outcomes? (i.e., lowest safe level of care appropriate to clinical need) 3. What are the efficiency implications of the pilot processes and outcomes? 4. What can be learned from the evaluation to inform future decisions? Evaluative criteria (attached at Appendix Two) were developed in consultation with the MoH, National Ambulance Sector Office (NASO), and the three providers involved in the pilots. A mixed method evaluation was undertaken, triangulating evidence collected through stakeholder interviews (7 funder and 19 provider individual or group interviews), a phone survey of callers (25 from each location) who were transferred to secondary triage, a clinical safety review of 9-10 calls in each location, and a desktop analysis of operational data. The evaluation found that the secondary triage pilots are being delivered well in each site. They are highly effective in terms of achieving their intended objectives (refer above). There is also a growing recognition, through the secondary triage pilot Evaluation of the secondary triage pilots in the Ambulance Communications Centres 1

5 process, of the ambulance sector as an integral part of primary health care, as well as part of the emergency response sector. More detailed findings are summarized below. We note that the evaluation is at a point in time and all provider organisations continue to make changes in the way the pilots are run and managed. We understand that some of the suggestions and opportunities identified for improvement now form part of the service delivered. Secondary triage processes Secondary triage is being delivered well by the providers in each location. All three providers have a high level of engagement and commitment at the strategic level and a big operational investment in making secondary triage work effectively, with a key person in each site overseeing and troubleshooting any issues. The evaluation highlighted very different approaches to doing secondary triage between the two locations. Medibank has a very systematic approach based on the progressive exclusion of serious concerns supported by algorithmic software. In Christchurch a diagnostic 1 or medical model of assessment is used that is based on pattern recognition. In terms of the balance between fidelity to a tool (clinicians adhering religiously to the script and guidelines) and clinical discretion, Medibank is located toward the fidelity end of continuum and St John the discretion end. (See diagram 1 on p.18.) Mostly both secondary triage approaches are working well, with some issues but real strengths in each place. In Wellington, Medibank provides well-established, systematic and robust secondary triage processes, using an internationally validated tool and robust governance and management processes. In Christchurch, St John is particularly good in ensuring the patient is able to access the agreed disposition, with a range of local community options based on good relationships and networks, and going the extra mile in their service to users. The evaluation s clinical safety review found there was not a consistent application of the secondary triage process or the triage tool by St John paramedics. In the Wellington calls selected for review, there was evident reluctance on the part of some nurses to refer back to WFA for transport. Whilst this may seem to support the aim of reducing numbers of ambulance transports, the clinical review team felt that it had the potential to affect access to health care for people with fewer financial and transport resources. Going forward with integrated healthcare design, the issue of transport only needs should be considered. The majority of callers transferred to secondary triage who are dealt with by hear and treat 2, are receiving a more rapid clinical response (in this case speaking with a 1 Both approaches involve seeking a provisional diagnosis, but one seeks to minimize or eliminate risk using systematic questioning before arriving at a preliminary diagnosis and plan (disposition) for the patient; whereas the other seeks to discover the cause of the chief complaint, (preliminary diagnosis) before going on to assess risks that may be associated with the emerging plan (disposition). 2 This service is perhaps more accurately described as hear and advise but hear and treat is the term used Evaluation of the secondary triage pilots in the Ambulance Communications Centres 2

6 nurse or paramedic) than they would have waiting in the code grey queue for an ambulance dispatch. Because secondary triage calls subsequently assessed as needing an ambulance response are returned to 111 with an upgraded priority code, these calls may receive a more immediate ambulance response than they would have without secondary triage. It was not possible to confirm this with data. Staff in all locations who were interviewed are supportive of secondary triage. Service user feedback is mixed, attributable in part to a lack of publicity, understanding, and conflicting messages between the ProQA script and the secondary triage process. Secondary triage outcomes The two secondary triage pilots were highly effective in achieving their intended objectives, to: Provide people calling 111 for an ambulance with the lowest safe level of care appropriate to their clinical need Contribute to the cost-effectiveness of that care, for example by a reduction in the number of lower acuity patients unnecessarily taken to an emergency department (ED) and improved utilization of current ambulance resources. Clinical safety standards have been met, and nearly 2000 calls resolved by secondary triage over 5 months. There has been a practically significant reduction in the dispatch of ambulances, freeing up resources equivalent to approximately one year s growth in demand. WFA transferred 2,713 calls to Medibank for secondary triage, and St John transferred 1,641 3 calls to their in-house secondary triage during the five-month period covered by the evaluation. Wellington resolved by hear and treat 54% of the total number of 111 calls transferred to secondary triage. Christchurch resolved 30% by hear and treat. (Confounding factors in interpreting this data include that WFA elected not to secondary triage some 111 calls aligned with a grey response while St John did so, and alarm calls are referred to secondary triage in Christchurch but not in Wellington.) Health efficiencies Savings from the secondary triage pilots include a decrease in ambulance dispatches equivalent to approximately one year s growth in demand. The estimated savings over the five months of the pilot on a marginal cost basis (the extra cost of staffing available ambulances) $109,000, and on an average cost basis (which takes into account the eventual need to add an extra crewed ambulance in response to growth) is $1.375 million 4. There are also improved health system efficiencies with the potential for further improvement. colloquially by both services, so is used throughout the report. 3 The substantial difference in numbers is due to the fact that the St John secondary triage pilot was operationalonly from 7am to 7pm, compared with Wellington operating 24/7. 4 If extrapolated to annual savings the translates to $262,000 based on marginal cost, or $3.3 million based on Evaluation of the secondary triage pilots in the Ambulance Communications Centres 3

7 Learnings to inform future decisions Success of the pilots has been demonstrated by the realisation of two key objectives: the lowest safe level of care appropriate to clinical need for users, and health system efficiencies. The pilots have provided proof of concept in relation to secondary triage. The following features of the pilots were identified as important contributors to their effectiveness and would need to be considered if secondary triage is scaled up: a 24/7 service and a quick response to secondary triage calls a seamless transfer of calls robust clinical governance and risk management processes a key manager in each provider organization who understands both business and secondary triage well a systematic approach to secondary triage, supported by good IT-based decision support training, support and ongoing staff development in telephone triage generally, and emergency medicine specifically a responsive secondary triage person who can balance safe outcomes for the user/patient (including enabling patients/users to access the recommended course of action) with operational imperatives, such as timeliness aligned data systems that enable analysis by and across the providers. All stakeholders consulted are in agreement that a single approach across the country would be ideal if secondary triage is scaled up. For a single approach to work well providers would need to reach agreement on: a single tool classification of determinants that can be effectively dealt with by hear and treat the transfer process used between 111 and secondary triage (warm transfer as used in Wellington or call-back as used in Christchurch 5 ) whether there is a preference about where secondary triage clinicians should be located 6. There was also agreement that roll-out of secondary triage needs an infrastructure in place to support it. This would include: a range of local health services that users can be referred to a range of transport options to enable users to access recommended actions from secondary triage The evaluation highlighted different approaches to secondary triage, on a continuum from a more process-oriented approach to a more diagnostic approach. At the average cost. 5 A warm transfer occurs when the call taker remains on the line until the caller is connected to the next person in the process. In a call-back process the caller is disconnected from the 111 call-taker and receives a call back from the next person in the process. 6 These possibilities were raised in the course of the evaluation. While they have not yet been seriously considered, they are one of a number of issues that will need to be addressed if a single approach to secondary triage is to be agreed. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 4

8 topline findings workshop there was consideration of the possibility that both approaches may be useful in particular circumstances. Feedback from the small group of service users surveyed indicated that satisfaction with a more processoriented approach was higher for cases that users considered to be of a less serious or urgent nature. This would suggest that a more diagnostic approach is better for more serious cases, but these ideas would need to be tested further. Investigation would also be needed to identify the ProQA determinants which may be better suited to resolution by hear and treat and are relatively easy to close off, and those which may require a more diagnostic approach. For any roll-out of secondary triage there would need to be some standardization of data capture and analysis across providers, and better processes for monitoring the health outcomes of people using secondary triage. For example, in the short to medium term there is a need to capture data on the number of re-presentations, where a second 111 call has been received after secondary triage. The single biggest issue that needs to be addressed if secondary triage is to be rolled out is user/patient understanding of what secondary triage involves, why they are being transferred, and the potential advantage of secondary triage for them. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 5

9 Introduction The Ministry of Health (MoH) contracted Evaluation Works Ltd to conduct an evaluation of the two secondary triage pilots being delivered, one in Canterbury and one in Wellington. This document presents the evaluation findings and identifies key issues for consideration going forward. Background The secondary triage pilots in the Ambulance Communications Centres (referred to as communication or call centres) are part of the Urgent and unplanned care integration project being led by MoH. This project is a national initiative focusing on unplanned care, where any person needs to access care quickly. The project aims to develop an integrated approach where people are consistently triaged and signposted to the care that is most appropriate to their needs. The secondary triage pilots have occurred over a period of continuous change in the ambulance sector. Clinical support desks in the communication centres were introduced in July in both Christchurch and Wellington, the new priority codes were introduced in August and the pilots just a month after this. The secondary triage pilots The pilots focus on 111 calls that are deemed low acuity and appropriate for secondary triage. Transfer of 111 calls to secondary triage means that, instead of an emergency ambulance being immediately dispatched, further clinical triage is provided for callers, and appropriate advice given on the health service they should access and in what timeframe. The range of options that could be provided includes: Referral back to 111 for dispatch of an emergency ambulance or dispatch of another health resource (e.g. acute demand ) Advice to urgently attend an Emergency Department Advice to see GP or an after hours clinic in 4 or 8 hours Advice to see GP within 24 hours Deferred referral pathway, for example to see GP within 72 hours Patient self-care. The objectives of the pilots are to: Provide people calling 111 for an ambulance with the lowest safe level of care appropriate to their clinical need Contribute to the cost-effectiveness of that care, for example by a reduction in the number of lower acuity patients unnecessarily taken to an emergency department (ED) and improved utilization of current ambulance resources. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 6

10 The pilots, which commenced on 1 October 2012, are: The SouthComms pilot, which covers the St. John Canterbury dispatch area (which is broadly consistent with the areas covered by Canterbury DHB) The CentralComms pilot, which covers the Wellington Free Ambulance greater Wellington region and Wairarapa dispatch area (which equate to areas covered by Capital and Coast, Hutt, and Wairarapa DHBs). There are some variations in responsibility for, and delivery of, secondary triage in the pilot locations. Strategic overview St. John is responsible for the SouthComms pilot. Secondary triage is a new component added to core business, and as such, is located in a single organisation with one culture. St John has developed a secondary triage tool (adapted from the internationally recognised Briggs nursing triage tool) specifically for use in the pilot, The triage tool was initially paper-based and then available on a tablet. From the outset it was recognized that a more robust system tool would be needed if rolled out further. The pilot was always viewed as a developmental, time-bound exercise to inform future decisions about secondary triage. Wellington Free Ambulance (WFA) in partnership with Medibank Health Solutions NZ Ltd (currently New Zealand s national Healthline provider) are responsible for the CentralComms pilot. Medibank handles the secondary triage component of the pilot, enabling each partner to concentrate on their core business albeit with some new processes within and between each organisation. Medibank uses an established, validated secondary triage tool. The algorithms are computer-based with a mature training programme and continuing quality assurance. The two organisations are located in different locations and have different cultures to be managed. The pilot was approached with a view to the potential to be ongoing and scalable. Delivery processes Secondary triage is used with 111 incidents that are coded grey 7 through the Medical Priority Dispatch System (MPDS) ProQA used in all communication centres. These incidents have been selected based on: The statistical probability of the patient being low acuity The MPDS determination that the patient has no immediate threat to life The likelihood that further evaluation might identify an alternative patient pathway (in some cases this may necessitate an urgent response) 8. In the Canterbury dispatch area, all such incidents are transferred to secondary triage; in the Wellington dispatch area, the majority of these are transferred to secondary triage, while some are referred to ambulance dispatch as usual. The The ambulance response system allocates 5 colour categories to determine the priority of emergency calls. Purple is immediately life threatening; Red is immediately life threatening or time critical; Orange is urgent and potentially serious but not immediately life threatening; Green is non-urgent (not serious or life threatening); Grey is also non-urgent (not serious or life threatening) with further telephone triage appropriate. 8 Wellington Free Ambulance and St John (2012). Secondary Triage Initiatives: Ambulance Communications Centres (COMMS) Education, Internal memo, 25 September 2012, pp 2,3. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 7

11 calls that both St John and WFA refer to secondary triage are coded grey. There is a small subset of ProQA determinants aligned to a grey response that St John refer to secondary triage, to which WFA allocate an ambulance response. These are coded gray. In the Canterbury dispatch area, all calls coded grey and gray during the 12 hour period from 7am to 7pm are transferred for secondary triage. These calls are handled by an advanced life-support (ALS) paramedic located at the clinical desk in the SouthComms call centre. A 111 caller whose condition is coded grey or gray is advised by the call-taker, after they have been taken through the ProQA script, that the call-taker is going to hang up and a paramedic will call them back within 30 minutes. The call-taker and the paramedic share information via the internal computer-aided dispatch (CAD) system. In the greater Wellington region and Wairarapa dispatch area, 111 calls coded grey are transferred for secondary triage. They are transferred to registered nurse call takers with Healthline directly via warm transfer (that is, the caller remains on the phone while WFA transfers key information to Healthline nurses 9 ). Healthline provides a 24/7 service. Process maps of the 111 caller s journey through each of the secondary triage processes are attached at Appendix One. Evaluation brief The evaluation covers the first five months of operation of the secondary triage pilot in SouthComms and Central Comms (also referred to as the Christchurch and Wellington pilots respectively), from 1 October 2012 to end February The evaluation covers set-up and implementation of the pilots, the day-to-day operation and preliminary, indicative outcomes. NASO and the MoH identified eight areas that they wished to see addressed in the evaluation of the pilots. These can be summarized as follows: 1. Achievement of the lowest safe level of care for the user/patient 2. Clinical safety in call handling, call disposition and clinical outcomes 3. The level of service user satisfaction 4. The level of acceptance by health professionals 5. Efficiencies achieved (such as the reduced use of ED and ambulances) 6. The impact of efficiencies on the wider health sector 7. Identification of key success factors in providing secondary triage 8. Anticipated effectiveness and value for money if rolled out. 9 Warm transfer is an operational precaution to ensure the service providers can confirm the transfer occurred, and the caller has made the connection with second call-taker. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 8

12 The stated use of the evaluation is to provide advice to NASO and the wider Ministry of Health on the following: what worked well, and what didn't, for the two pilots, and whether the outcomes are better/worse/the same as the status quo whether one or both pilot(s) should be continued and/or extended and in what format any service provision, service user safety, cost implications of expanding either pilot option across New Zealand. As the secondary triage pilots are part of the Urgent and unplanned care integration project, and link to other Ministry-led integration projects at the national level, there is also interest in the extent to which the pilots, by changing service delivery arrangements, contribute to achieving more integrated care. Methodology Key evaluation questions (KEQs) Four key evaluation questions 10 were identified to focus the evaluation. These are the overarching questions that the evaluation aimed to address: KEQ 1: How well are the secondary triage processes working in each pilot site? KEQ 2: How well are the secondary triage pilots achieving their intended outcomes? (i.e., lowest safe level of care appropriate to clinical need) KEQ 3: What are the efficiency implications of the pilot processes and outcomes? KEQ 4: What can be learned from the evaluation to inform future decisions? Evaluation design To ensure the evaluation would produce robust, credible findings of practical value to NASO, MoH and providers, an evaluation-specific methodology was used to answer these questions. Central to this approach was the development of evaluative rubrics (see Appendix Two) to enable clear judgments to be made about how good the pilots were, how they could be improved, and which aspects might be continued or rolled out. The rubrics make transparent the criteria by which the pilots were evaluated, setting out clear descriptions of performance from highly effective to ineffective, and a range of levels in between. The criteria were developed at the outset in consultation with NASO and the Ministry of Health, and other stakeholders 11 during a full-day workshop. 10 The KEQs were tested with stakeholders at the rubric workshop on 7 February. 11 Stakeholders in addition to NASO and MoH included St John, Wellington Free Ambulance, and Medibank Health Solutions. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 9

13 Alongside the rubrics were some exploratory questions aimed at understanding patterns in the level of care to which people were triaged, and efficiencies resulting for the pilots, as well as identifying possible future directions for secondary triage. The evaluation employed a mixed-methods approach, drawing on both quantitative and qualitative data, to enhance the validity of evaluative conclusions by enabling corroboration of findings from multiple sources. Four data sources were used: face-to-face interviews with funders and providers (paramedics, Healthline nurses and communications centre staff) a small telephone survey with service users a clinical safety listening to calls review of a small sample of pre-recorded calls a desktop analysis of operational data and financial data. Analysis of the findings are organized in relation to the KEQs. Methods The following sections specify the methods used to address the key evaluation questions. Stakeholder interviews Stakeholder engagement included: Preliminary interviews with 5 NASO and MoH staff, 2 senior staff each from St John and Wellington Free Ambulance, and 3 from Medibank in Wellington Familiarisation visits to each of the providers - St John in Christchurch, Wellington Free Ambulance and Medibank including meeting communications centre staff Individual and group interviews with providers, including those managing the pilots in each site, paramedics and nurses undertaking secondary triage, operational paramedics, and communications centre staff including: o 3 interviews with managers of the secondary triage pilots o 5 individual interviews with nurses/secondary triage paramedics o 3 group interviews with operational paramedics o 1 group interviews and 3 individual interviews with comms centre staff 12. The assistance of all three providers in both facilitating introductions and making arrangements for visits is gratefully acknowledged. 12 Selection of those interviewed, apart from the interviews with the three managers, occurred as follows. The evaluators identified the range of people they wished to interview (e.g. clinical desk paramedics, Healthline nurses, comms centre call-takers) and whether by individual or group interview, and asked providers to arrange this. Because all providers are heavily operational, what this often meant in practice, was that the evaluators interviewed those who were on shift on the day and available, at the time the evaluators were there. This reality reduced the likelihood of bias, and evaluators were impressed by the openness and frankness of interviewees. No-one refused to participate. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 10

14 It had originally been envisaged that a wider group of health professionals (including staff at Emergency Departments) would be interviewed, but this was not considered feasible as they were unlikely to have knowledge of the pilots. Telephone survey to service users The phone survey gathered early feedback from callers (25 from each location) who had been transferred to secondary triage 13. The focus of the phone survey was users understanding and satisfaction with the secondary triage service, about what worked well and less well, the outcomes they experienced, and issues they felt should be addressed. The survey was exploratory, and intended to provide an indication of the range of responses and experiences that users had (in the context of an absence of information to the public about the secondary triage pilots) and to highlight potential issues that may need to be addressed if secondary triage is to be rolled out. The callers were randomly selected from a spreadsheet of anonymised secondary triage service users over a two-week period from February Initial planning to ensure coverage of a range of demographic factors (gender, ethnicity, age) and both medical and trauma conditions, was unable to be met because of different data sets available in the two locations 14 and difficulties contacting people, despite calls being made at different times throughout the day (including the evening) and during both weekdays and weekends) 15. The first 25 people who were available to respond to the survey were selected 16. It became apparent in conducting the survey that a number of respondents had a compromised recollection of events. In some cases this related to the stress of the situation, and in the case of Christchurch, respondents were aware they were talking to St John but didn t differentiate between 111 and secondary triage. For those who subsequently received an ambulance response, it was difficult to separate out satisfaction with the ambulance from satisfaction with the secondary triage process. In addition just as the survey was being implemented, the evaluators became aware that the ProQA script read to 111 callers in Canterbury 17 prior to being transferred to secondary triage tells them that a paramedic or ambulance is on its way, and in a small number of situations this also occurs in Wellington, resulting in confusion. All of these factors contribute to a need to treat the survey findings with caution. The phone survey is attached at Appendix Three. 13 The number of survey respondents was determined by the MoH budget available for the evaluation. 14 In one location, ethnicity data was not captured (although was subsequently gathered as part of the survey), and in the other, the detailed condition was not summarised as being a medical or trauma condition. 15 The survey was intended to be exploratory rather than generalizable so representativeness was not mandatory. As noted above the survey findings are indicative only, but where there is a high degree of consensus are highly suggestive of a shared issue, or when considered with other data, can provide a useful triangulation. 16 Information about the respondent population is discussed in KEQ 1 under the heading Credibility and acceptability to service users. 17 This issue has since been rectified. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 11

15 Clinical safety review of calls This aspect of the evaluation focused on clinical safety during the secondary triage call. The evaluation team s clinical team 18 retrospectively listened to 9 secondary triage calls handled by Medibank nurses, and 10 calls from St John 19 when the paramedic on the clinical desk undertook secondary triage. The focus of the review was to: assess fidelity with the clinical support decision-making software gather evidence of appropriate clinical questioning comment on any potential safety issues arising. The calls were randomly selected 20 from a spreadsheet of anonymised secondary triage service users over a two-week period from February These were a different group of calls than those selected for the phone survey. Desktop data analysis The desktop analysis focused on operational data relating to clinical outcomes, and financial data relating to efficiency and/or effectiveness. Operational data was analysed from both of the pilot locations against a common set of indicators including: operational indicators (such as numbers of 111 callers referred to secondary triage, or the number of times particular triage options were used) clinical indicators to ensure clinically safe outcomes (such as defined level of acuity pre- and post secondary triage). Analysis of financial data is used to inform a discussion about the value for money of the secondary triage pilots. There are opportunities to improve alignment of data gathered by the different providers, in order to better aggregate information about secondary triage and to compare and contrast operational and outcome data. The issues and limitations with interpreting, comparing and contrasting the data from the two pilots are identified in the Findings section. Analysis Analysis and reporting of the pilots occurs at both the individual pilot level, and across pilots, and is organized in relation to the key evaluation questions. This means that each pilot has been considered against the evaluative rubrics and exploratory questions, so that judgements against previously-determined ratings (such as ineffective, minimally effective, developing effectiveness, consolidating effectiveness, and highly effective) can be made. 18 The clinical team comprised a General Practitioner, an Emergency Department nurse with 20+ years experience, an experienced paramedic/paramedic educator, as well as the evaluation team s project leader. The composition of this clinical team was proposed and approved by MoH. 19 MoH contracted the evaluators to listen to between 6-10 calls, and these were the numbers completed in the equivalent amount of time set aside for each site. 20 In an attempt to make the calls as representative as possible. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 12

16 A value for money discussion is focused on the efficiency implications of the pilot processes and outcomes (see KEQ 3), and the associated exploratory questions (see Appendix Two) define the scope of this analysis. Findings The secondary triage pilots overall meet the criteria for developing effectiveness (as defined in the evaluative rubrics) in relation to how well secondary triage processes are working, and in Wellington, this is close to consolidating effectiveness. Both are highly effective in achieving intended outcomes, but in the case of Christchurch, this is qualified by the fact that these results are heavily influenced by a default to a see and treat response. The Wellington pilot was more effective in reducing ambulance call-out. There have been clear efficiencies in ambulance utilization in both locations, and the pilots have identified key success factors for informing future decisions. The following paragraphs present the evidence that supports these findings. KEQ 1: How well are the secondary triage processes working in each pilot site? Overall finding (level of effectiveness against rubric 1) Secondary triage is being delivered well by the providers in each location. All three providers have a high level of engagement and commitment at the strategic level and a big operational investment in making secondary triage work effectively, with a key person in each site overseeing and troubleshooting any issues. Mostly secondary triage processes are working well, with real strengths and some issues in each place. In Wellington, Medibank provides well-established, systematic and robust secondary triage processes, using an internationally validated tool and robust governance and management processes. In Christchurch, St John is particularly good in ensuring the patient is able to access the agreed disposition, with a range of local community options based on good relationships and networks, and going the extra mile in their service to users. Staff in all locations are supportive of secondary triage. Service user feedback is mixed, with both positive and negative responses. The following is the list of factors that stakeholders agreed mattered in terms of a good secondary triage process: triage tool range of dispositions recruitment, training and selection of staff Evaluation of the secondary triage pilots in the Ambulance Communications Centres 13

17 clinical governance and risk management clinical safety seamless interface timeliness credibility and acceptability service users and health professionals mechanisms for learning and improving. Each of these is discussed below. Triage tools Different secondary triage tools are in use in the 2 pilots. The CareEnhance Call Centre (CECC) secondary triage tool used in Wellington has established credibility internationally, based on experience of utilisation and research to validate its effectiveness. Currency and fit of the tool are optimised with processes for localization, first to Australia then to NZ, and continuous development. CECC has credibility to the nurses and medical staff in Medibank in terms of its application in the local pilot context. The Clinician Applied Secondary Triage (CAST) tool used in Christchurch has been adapted from the Briggs Telephone Triage Protocols for Nurses by St John, with signoff from Craig Ellis the Medical Advisor. While the Briggs tool has established credibility and validity in an overseas context, it has not been validated in its current adapted form. There are no concerns about risks to safety based on evidence from the clinical audit, observations by medical advisors and an absence of adverse event but stakeholders did not see it as a tool to take to scale. The tool was selected for the pilot on the basis of cost and speed of implementation, and organizational preference for any future rollout would be for a fully comprehensive paramedic or clinical tool. Range of dispositions Dispositions were thought about in different ways in the two locations. In Christchurch, secondary triage staff tended not to strongly differentiate between a set of secondary triage dispositions and usual ambulance dispositions, rather seeing the dispositions available to them as operating on a continuum from self-care to sierra cars 21 to emergency ambulance dispatch. In Wellington, on the other hand, where hear and treat (secondary triage) and see and treat options are dealt with by separate organisations, there is more of a focus on the discrete secondary triage disposition options. 21 Sierra cars are a new emergency response vehicle being used by St John. The main aim of the sierra cars is to reduce lower acuity work and where possible work with patients in the community to ensure the best clinical pathway is followed. Sierra cars are equipped with everything a frontline ambulance would carry, apart from stretchers and back-boards. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 14

18 Staff in both locations reported being satisfied with the number and range of dispositions available to them during the pilots, although it was suggested that other options might usefully be made available if secondary triage is to be scaled up. These include referral to a pharmacy and other community-based dispositions. The second aspect in relation to dispositions, is the extent to which the secondary triage nurse or paramedic does or should help the person to access the determined disposition. Localised support appears to be an important factor in this. Christchurch has a good network of local support options and proactively facilitates access as part of a strong people-centred culture focused on the patient getting the appropriate health care. In Wellington this support was variable. Relationship managers are very clear that it is not the role of Medibank to be a gatekeeper of access to ambulances for transport. However, interviews with a small number of Medibank nurses, feedback from survey respondents, and from the clinical team, suggest that some Healthline nurses appear to be setting a high bar for the patient before ambulance transport will be requested. There was wide agreement amongst stakeholders that access to a wide range of transport options was required to support people to access the recommended action following secondary triage. Recruitment, selection and training of personnel Secondary triage staff received an acceptable minimum level of training in the application of the secondary triage tools and call transfer processes (to and from the communications centres). In Christchurch paramedics would value more in-depth training in telephone triage and the application of the secondary triage tool, and staff at WFA considered that Medibank nurses could usefully benefit from knowing more about ambulance codes when referring calls back. Some nurses suggested that specific training in triaging emergency situations would be helpful. Both locations had ongoing secondary triage staff development processes (e.g., continuous quality improvement and experiential learning) in place, although those at St John were on a more as needs basis. Medibank had more robust processes in place for supporting staff in a systematic way e.g. listening to 1% of calls for each nurse each month, and providing targeted staff development and support as necessary. Their processes contribute meaningfully to improvement over time. Call takers in the two pilot communications centres generally felt adequately supported during the introduction and development of secondary triage, even though it was a very stressful time given the speed of its introduction, and that it occurred at the same time as the introduction of priority codes and the clinical desks. Clinical governance and risk management All three providers have pre-existing robust clinical governance and risk management frameworks and processes that are now applied to secondary triage. These include processes to deal with any adverse events or near misses. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 15

19 Each provider developed new processes to support secondary triage. In Wellington, these processes were systematic and ongoing at both WFA and Medibank. For example, WFA closely monitored the ProQA interface with secondary triage calls and Medibank, from the outset, monitored all calls from 111. Both Medibank and WFA built in a process of reviewing all calls that were returned to 111. Processes are sufficiently aligned with each other to support effective clinical governance. In Christchurch, new processes were developed for the purpose of a three-month trial. This meant they were developed as one-off processes rather than integrated into current systems. The focus has been on doing what is valid and useful within the 3 months pilot period. For example, one clinical safety audit was conducted over this time. At the end of the evaluation period, St John had identified opportunities to improve safety and quality but were waiting on a decision re pilots before acting on them. In Wellington secondary triage staff received supervision and support that, at a minimum, includes a continuous process of monitoring and reviewing calls, and reviewing staff performance. In Christchurch, at the time of the evaluation, there was no systematic supervision and support after initial training and no systematic process of monitoring and reviewing secondary triage calls. A range of ad hoc support and monitoring including weekly conferences and listening to randomly selected calls was in place. Medibank has processes for aligning the CECC triage tool with new international and local developments and learnings from nurses. In Christchurch, the development of the CAST tool, and its use, were subject to usual clinical governance and risk management frameworks. Both WFA and St John track international developments in secondary triage. Clinical safety This section focuses on application of and fidelity with triage tools and processes, clinical questioning, reasoning and decision-making. In both locations secondary triage staff safely distinguish between calls that are safe to hear and treat, calls where there is sufficient information to hear and treat, and calls that require a see and treat response. The evaluation highlighted very different approaches to doing secondary triage between the two locations (as illustrated in the diagram below). Medibank has a very systematic approach based on the progressive exclusion of serious concerns supported by algorithmic software. In Christchurch a diagnostic 22 or medical model 22 Both approaches involve seeking a provisional diagnosis, but one seeks to minimize or eliminate risk using systematic questioning before arriving at a preliminary diagnosis and plan (disposition) for the patient; whereas the other seeks to discover the cause of the chief complaint, (preliminary diagnosis) before going on to assess Evaluation of the secondary triage pilots in the Ambulance Communications Centres 16

20 of assessment is used that is based on pattern recognition, listening for cues and using mini-deductive processes in a free-form way. This aligns more with the approach used by paramedics during a see and treat response, or what a general practitioner might use. In terms of the balance between fidelity to tool and clinical discretion Medibank is located toward the fidelity end of the continuum and St John the discretion end 23. Diagram 1: Approaches to secondary triage Nurses at Medibank apply the CECC tool in line with the guidelines. The evaluation s clinical review of a small number of calls concluded that in most cases the nurses effectively balanced fidelity to the clinical decision support tools and clinical discretion, and elicited necessary and sufficient information through questioning to support a valid clinical decision. In some cases, the clinical team queried the sufficiency of the information and the speed with which a clinical decision was reached. St John paramedics, are not consistently using the CAST tool in line with the guidelines, rather as a reference or check on clinical reasoning. This approach relies very heavily on the clinical questioning and reasoning skills of the individual practitioner, so requires a high degree of quality control, which an internal audit concluded was present. The audit also recommended more intensive oversight of secondary triage in the form of increased medical advisor hours. The more substantive issue for St John that emerged during the evaluation s clinical review was the apparently ad hoc approach to secondary triage during the call-back risks that may be associated with the emerging plan (disposition). 23 Neither Medibank or St John felt that the descriptors ( progressive exclusion of serious concerns and diagnostic ) were quite right, but there was general understanding about what these labels were attempting to describe at the topline findings workshop with stakeholders. St John subsequently suggested labels of prescriptive and adaptive as alternatives but there has been no opportunity to test these. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 17

21 made by the clinical desk paramedic. Of the calls listened to, in some cases secondary triage was not done at all, with the return phone call more a reassuring conversation. When secondary triage was done, sometimes the information gathered was insufficient and sometimes there appeared to be no reference to the tool. Comments from St John in response to this analysis note that they do not see secondary triage as a discrete process that occurs once the call-back occurs; rather it begins when the incident hits the waiting incident queue. Calls can be upgraded based on a notes review, followed by a phone call from the clinician to the patient reassuring them that help is on the way. The clinical team found that there was no patient safety risk. The team s view was that, in large part, this was due to a tendency to default to see and treat. Strictly speaking, this does not provide evidence of the clinical safety of the use of the hear and treat option, as the telephone triage process has decided in favour of an ambulance response. St John would propose a holistic view of the triage process, beginning from the moment the incident hits the waiting incident queue. At any time a red flag in the incident notes or the questioning process can generate a see and treat response. This means that some calls are upgraded based on notes review and a phone call will follow from the clinician to the patient reassuring that help is on the way. This underpins patient safety in the overall process. One issue that emerged related to access to transport for those users referred to ED or their GP immediately in the Wellington area. The clinical safety review and two other evaluation methods found that the reluctance of some nurses to refer callers back to WFA for transport 24, was a source of stress to users. The nurse seemed angry with me because I had no way of getting to hospital 25. Whilst this may seem to support the aim of reducing numbers of ambulance transports, the clinical review team felt that it had the potential to affect access to health care for people with fewer financial and transport resources. Seamless interface Within the pilot regions, there is generally safe transfer between the ambulance communications centres and secondary triage. There is minimal need for patients to repeat information at handover points, and no critical information is lost. There are mechanisms in place to recover lost calls, or default to an ambulance if the call can t 24 The evaluation team was informed by Medibank staff that, as part of the secondary triage pilot, callers whose disposition was to go to ED or their GP immediately, and who didn t have transport, should be referred back to 111, and that the bar for this was expected to be low during the pilot. 25 Feedback from one of the survey respondents. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 18

22 be retrieved. Handover now takes place in a smooth and seamless manner to the satisfaction of clinicians, virtually all of the time 26. Calls from out-of-region communication centres are often not appropriately transferred to secondary triage in the pilot areas. In part this is due to the complexity of the different processes to be applied for Christchurch and Wellington in respect to grey or gray codes, and partly to NorthComms call centre managers determining at a particular point in time that it was not safe to overload their 111 call-takers with further changes in secondary triage protocols, given other changes that they were integrating. There have been some problems associated with ProQA scripting that have impacted on the interface between 111 call-takers and secondary triage staff in both locations. ProQA requires particular scripts to be read out to 111 callers by call-takers, and this meant that callers had already been advised an ambulance was on its way before being transferred to secondary triage. The information read out to 111 callers as per the ProQA script, advises callers at the outset that I am organizing the paramedics to help you now. Unsurprisingly callers were confused about what was happening, especially in Wellington where they are transferred to a different organisation, and nurses at Medibank nurses initially had to deal with some hostile reactions. Problems with scripting have been, and are being, progressively addressed by providers to enable secondary triage processes to work effectively. Some delay has occurred due to having to negotiate with the ProQA script owner. Wellington appears to have been able to address these more swiftly. Timeliness The majority of callers transferred to secondary triage who are dealt with by hear and treat are receiving a more rapid clinical response than they would have if an ambulance was dispatched. Existing timeliness standards, as specified in current contracts, have continued to be met, and the timeliness of the initial response, secondary triage and return to 111 when needed, is considered safe by the providers. Virtually all calls coded grey in the WFA region were immediately transferred via a warm transfer process to the Healthline nurse. Grey calls handled by other Communication centres received a variable response, with some receiving secondary triage and others not. This was due to the complexity of the different secondary triage processes for each pilot, the virtual call-taking environment and transfer of calls to other centres, and work overload issues in NorthComms. Due to separate IT systems in the two organisations, it is not currently possible to provide data regarding the average time a secondary triage call took from being received in the communications centre to resolution. 26 In this evaluation: majority means more than half; vast majority means around three quarters or more; virtually all means close to 100% allowing for reasonable exceptions. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 19

23 Some of the service users surveyed from the WFA region, who received secondary triage via Medibank and were returned to 111, did express concern about the time taken to respond to a further set of questions when they felt action was needed immediately, or did not feel well enough or in a position to respond. This experience possibly reflects the absence of a public awareness campaign and lack of understanding that secondary triage can accelerate an ambulance response for calls returned to In Christchurch, 111 calls coded grey and gray and transferred to secondary triage received a similar response time, compared with similar types of calls received over the same 5 month period a year earlier (October February 2012). The timeliness standard instituted by St John for the secondary triage process was based on a call-back from the clinical desk paramedic within 30 minutes, or if that was not possible, by the original call-taker to check any change in symptoms. If there was no change then the same process was repeated for a further 30 minutes. If there was a change for the worse, or the caller could not be contacted, the call was escalated. If the paramedic was not able to respond within 60 minutes, the call was upgraded and an ambulance dispatched. St John s timeframes for the secondary triage processes were met. All secondary triage calls (whether resolved by hear and treat or by ambulance dispatch) were, on average, responded to within the 30 minute call back period. For grey and gray calls resolved by hear and treat, the average time taken from the time the 111 call was received to resolution was 24 minutes For those calls resulting in an ambulance dispatch, the average time it took from the time the call was received, for the paramedic to undertake hear and treat, upgrade the call and for an ambulance to be dispatched was 21 minutes. As in the Wellington pilot, those secondary triage calls that subsequently require an ambulance response may receive a more immediate response than they would have without secondary triage 28. No concerns were raised by the surveyed service users in Canterbury about timeliness in relation to secondary triage. However this needs to be interpreted with caution as many 111 callers in Christchurch were not fully aware that something different from the usual ambulance response was happening, as the secondary triage call was from a paramedic from within the ambulance service, and in most cases they had also been told an ambulance was coming. 27 Grey calls returned to 111 may receive a more immediate ambulance response than they would have without secondary triage due to being upgraded in terms of response priority. This was unable to be verified with data. 28 Refer footnote 27. Response times were unable to be compared with a similar period a year earlier given that the priority coding system changed just prior to the introduction of secondary triage pilot, and all calls prior to secondary triage resulted in an ambulance dispatch. Further investigation of current timeliness data may be possible with St John but due to the complexities in programming, analysis and interpretation was beyond the scope of this evaluation. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 20

24 Credibility and acceptability to service users There are three important factors that provide context to the findings from the survey of fifty 111 callers: 1. There was no publicity about the introduction of the secondary triage pilots, so 111 callers were often confused and had no understanding about what was happening. A number of phone survey respondents found it difficult to separate out the secondary triage aspect from the entire 111 call process. 2. As noted in the earlier section Seamless interface, the information read out as per the ProQA script gave callers to understand that an ambulance was on its way. This information was provided to all 111 callers in Christchurch and some in Wellington. 3. In addition, in Christchurch, an ambulance was sometimes dispatched prior to the 111 caller being transferred to secondary triage. In total 20 of 25 Christchurch callers received a see and treat response. Fifteen of 25 Wellington callers were referred back for an ambulance. These factors mean that it is difficult to interpret differences in service user experience and satisfaction between the two different pilots. Comments in the survey by people who subsequently received an ambulance, suggest that levels of satisfaction are often correlated with their satisfaction at receiving an ambulance rather than secondary triage. The information that follows is drawn from the survey of 25 people from each location, who had been referred to secondary triage after calling 111 during February In each site, 18 or 72% of the respondents were female and 7 or 28% male. In terms of age, the Wellington/Wairarapa population was a broader crosssection than the Canterbury population, with the respondents spread pretty evenly across all categories. Eight were under 40, 7 aged between 40 and 60, and 10 being older than 60. In contrast, there were 5 Canterbury respondents under 40, 2 aged between 40 and 60, and 18 being older than In terms of ethnicity, the vast majority (44 or 88%) of respondents were New Zealand European/Pākeha, with another 4 respondents (8%) identifying as European (no further definition). The remaining two respondents were Indian and Samoan respectively (both from Wellington/Wairarapa). Just over half of the respondents (27) called the ambulance for themselves and the remainder (23) for someone else. Sixty percent (15) of the callers in Christchurch and 36% (9) of callers in Wellington understood why they were being referred for secondary triage. More than 50% of callers in both pilots were positive about being transferred to secondary triage. More people in Christchurch were very positive than in Wellington 29 There are two factors that may be affecting this data that St John included alarm calls in their secondary triage data and Wellington did not, and that falls are included in the grey calls that St John responds to, and are not included by Medibank. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 21

25 (9 compared with 5) and less people in Christchurch were negative than in Wellington (2 compared with 11 people). The vast majority 30 of people surveyed were satisfied with the secondary triage service in Chch (23/25) compared with the majority in Wellington (17/25). A third of people in Wellington were unsatisfied whereas just 1 person in Christchurch said they were unsatisfied. There were differences between pilots in the open-ended survey responses about what worked well. Almost half of the Christchurch users used the word reassuring or comforting in describing the best things about their experience of the secondary triage process. Some of the comments suggest that there was a level of comfort in talking to a paramedic who is located within the ambulance service that they trust. In contrast, comments in relation to Medibank nurses tended to focus positively on their efficiency and competence (7/25) 31. Key points of dissatisfaction were users feeling confused and stressed, the length of time that secondary triage took, not knowing if ambulance coming, the nurse s attitude or comments, and needing to find their own transport. There also seemed to be a positive correlation between high satisfaction and relatively lower levels of acuity of a condition (as perceived by the user), and vice versa. Just didn t like the process although can understand it could be good where a case wasn t urgent. His case was urgent, he needed instant sedation when admitted, and remained in hospital for 3 weeks If things hadn t settled down, it would have been difficult if no-one was there to help and you were talking to someone on the phone while the child was choking The nurse didn t listen to what I was saying, when I knew there was an issue. In the end I just answered the questions as quickly as possible to get through that part. 32 This suggests an openness by users to secondary triage in low acuity cases. Good system to have in place for non-emergencies In this evaluation: majority means more than half; vast majority means around three quarters or more; virtually all means close to 100% allowing for reasonable exceptions. 31 Four users focused on the skill base of paramedics in Christchurch and four users in Wellington also described the nurse as reassuring. 32 All 3 quotes are feedback from survey respondents. 33 Ibid. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 22

26 Credibility and acceptability to health professionals The secondary triage pilots have high credibility with, and acceptability to, the paramedics, Medibank nurses, and communications centre staff working for the provider organizations who were interviewed. There is good sector buy-in. This is based on beliefs about secondary triage addressing low acuity calls 24/7 being the right thing for the user, as well as ensuring better utilization of resources, including ambulances and staff time. Paramedics reported having noticed a difference and that they were now able to have meal breaks, which is a huge issue for working paramedics. Amongst this group of staff there are variations in knowledge and level of direct involvement with secondary triage, for example WFA paramedics had little awareness about the pilots beyond that they were happening. Other staff are heavily involved at the front line, for example, Medibank nurses who had to deal with some angry callers in the early days, and in actions to streamline and finesse the process. There are ongoing discussions for example, about whether the right calls are being referred to secondary triage, that is, whether the ProQA determinants currently coded grey or gray are all low acuity and appropriate for hear and treat. It is anticipated that such discussions will be ongoing as secondary triage is being refined. Another concern raised by the clinical review team was whether secondary triage may inhibit access to ambulance transport, and therefore the recommended level of health care, for those with limited resources 34. There is a growing recognition, through the secondary triage pilot process, of the ambulance sector as an integral part of primary health care, as well as part of the emergency response sector. In this way the pilots can be seen to be facilitating health sector integration. Mechanisms for learning and improving Central to the process of learning and improving in relation to the secondary triage pilots has been the presence of a key person or persons in each of the three provider organisations 35, who have maintained constant oversight of the project, and have developed and monitored the necessary mechanisms for learning and improving. These people have been able to hold a wider view of the organisational objectives for secondary triage as well as attend to the operational issues. In both locations, where operational issues arise or problems have emerged there are mechanisms to ensure learning occurs and improvements are made to the pilots as a result. WFA and Medibank each developed new systems to support secondary triage. For example, WFA closely monitored the ProQA interface with secondary triage calls and Medibank, from the outset, monitored all calls from 111. Both 34 The evaluation team considers this to be a risk that MoH needs to be mindful of in any possible roll-out. 35 In Medibank Janet Harp and Anne O Brien, at St John initially Sue Gullery and now Dan Ohs, and at WFA initially Andy Long and now Paul Fake and Ruth Lloyd. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 23

27 Healthline and WFA built in a process of reviewing all calls that were returned to 111. St John developed one-off processes for the pilot rather than integrated these into current systems e.g. carried out one clinical safety audit. Medibank has effective strategies and processes for identifying and documenting reflective learning that is deliberately focused on future implementation of secondary triage. In Christchurch, secondary triage staff received ad hoc support and monitoring including weekly conferences and listening to random calls. There was no systematic supervision and support after the initial secondary triage training. St John have identified opportunities to improve safety and quality in their secondary triage approach which have since been implemented. St John, WFA and Medibank all capture relevant and useful data that is actively being used to inform their respective ongoing service enhancements. The following section highlights a number of limitations in comparing and contrasting the data. Improvements in the alignment of data gathered by the different providers is recommended, in order to better aggregate information about secondary triage and to compare and contrast operational and outcome data. KEQ 2: How well are the secondary triage pilots achieving their intended outcomes? Overall finding (level of effectiveness against rubric 2) The two secondary triage pilots were highly effective in achieving their intended outcomes. The pilots are delivering on the primary clinical objective, which is to achieve the lowest safe level of care appropriate to clinical need. Clinical safety standards have been met, and nearly 2000 calls resolved by secondary triage over 5 months. There has been a significant reduction in the dispatch of ambulances, freeing up resources that represents approximately one year s growth in demand. The following factors are those determined in the rubric to be key to achieving intended outcomes: Clinical safety Level of care to which people are triaged Impact on ambulance dispatch Impact on emergency departments. Clinical safety The primary clinical outcome for the secondary triage pilots is the lowest safe level of care appropriate to clinical need. By definition, 111 calls transferred to secondary triage are receiving a lower level of care than an immediate transfer to ED. Clinical Evaluation of the secondary triage pilots in the Ambulance Communications Centres 24

28 safety is concerned with ensuring that secondary triage addresses the lowest level of care in a way that is safe for the user/patient and appropriate to need. In terms of safety, the pilots are delivering on this objective. In both locations, virtually all 36 calls that received a clinical audit 37 were found to be safe. Of 84 cases audited at St John, there was a single case of incorrect advice or care 38. There were no adverse events (or near misses) identified by Medibank or St John 39. Based on the small number 40 of secondary triage calls listened to retrospectively by the clinicians on the evaluation team, all calls listened to in each site were judged to be clinically safe with no serious concerns about safety or inappropriate dispositions. This judgement was based on different reasons. In Wellington there was evidence of robust triage processes and systems, and for the most part, evidence of sound clinical reasoning. However, in the case of St John, secondary triage formed part of their wider service provision as opposed to a discrete service. The confidence that the calls were clinically safe was due to either an ambulance being dispatched as well as referring the call to see and treat, or a tendency by the clinical desk paramedic to default to see and treat, rather than because the secondary triage process as a discrete activity, was proven to be safe. In effect, secondary triage per se was less rigorously tested at St John during the pilot 41. As noted previously (see p.12), there is an associated non-clinical safety issue that emerged in relation to access to transport for some users referred to ED or their GP immediately in the Wellington area. Three information sources in the evaluation to indicated that some nurses acted as gatekeepers to these users being referred back to WFA for transport. As well as being a stress to users this raises a question about whether these actions prevent users from accessing recommended health care. Based on the clinical audits, the clinical review undertaken by the evaluation team, and absence of adverse events or near misses, the dispositions determined by nurses and paramedics were appropriate to need. A more robust way of measuring appropriateness or sufficiency of care however, would be to capture data on the number of re-presentations, where a second 111 call has been received after secondary triage (as it is possible that secondary triage could result in high resolution rates in the short term, but that callers have to return). This data is not currently available, but could probably be captured in the future. 36 Close to 100%. 37 All three providers regularly undertake clinical audits of their calls, as per their contracts with MoH and as part of their own clinical safety processes. 38 Comparable data was not sought from Medibank. 39 One survey respondent identified a near miss, but upon investigation this was found to be not a secondary triage issue as the caller had been referred back to 111 by Medibank. 40 Nine in Wellington and 10 in Christchurch. 41 This may in part be due to the way that the clinical reasoning behind a decision was not articulated during the calls that were reviewed, and therefore were not able to be evaluated. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 25

29 Level of care to which people are triaged The following provides a description for each of the pilots, of the total call volume and acuity, proportion of calls eligible for secondary triage, resolution of secondary calls and the level of care to which calls resolved by hear and treat were triaged. Over the period of the pilot (Oct 2012 Feb 2013), the two providers received the following number of 111 calls: Table 1: Total number of 111 calls 42 Christchurch pilot Wellington pilot 12,317 23,188 The substantive difference in call volumes reflects the different operational hours between the two pilots. WFA and Medibank provided secondary triage 24/7, and St John offered secondary triage over a 12-hour period from 7 am to 7 pm when the clinical desk was staffed by a paramedic. About one-fifth of the total number of calls in each of the pilot areas were coded grey/gray 43. Chart 1: Distribution of 111 calls over response codes 39% 37% 32% 28% 20% 21% 8% 13% 1% 1% Purple Red Orange Green Grey/Gray St John Wellington Free Ambulance 42 The two providers calculate the total number of calls differently. St John counts the number of discrete incidents and excludes patient transfers (use of ambulance for non-emergency transport), whereas WFA counts all calls irrespective of whether they concern the same incident. This would suggest that if WFA used the same basis for calculating the total number of calls, the percentage figure regarding resolution of 111 calls by secondary triage would be higher. 43 In the Canterbury dispatch area, all incidents coded grey are transferred to secondary triage; in the Wellington dispatch area, the majority of these are transferred to secondary triage, while some are referred to ambulance dispatch as usual. The 111 calls that both St John and WFA refer to secondary triage are coded grey ; the small number of ProQA determinants aligned to a grey response that St John refer to secondary triage, and WFA to ambulance response, are coded gray. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 26

30 Of the calls coded grey/gray, those that were considered eligible for (and transferred to) secondary triage reduced to 12-13% of the total number of calls to each provider 44. Chart 2: Referral to secondary triage (as percentage of total number of 111 calls) 87% 88% 83% 13% 12% 17% Eligible Not eligible St John Wellington Free (actual) Wellington Free (theoretical) WFA assessed that a further 5% of the grey calls were also theoretically eligible for secondary triage 45. Just under one-third (30%) of St John, and over half (54%) of Wellington Free Ambulance referrals to secondary triage were resolved by hear and treat The difference between the initial grey coding (20%) and subsequent eligible grey calls (13%) for St John is due to grey calls which occurred outside of clinical desk hours (7am 7 pm), where the secondary triage process was not followed by the call taker (most commonly by those located in other call centres) or the clinical desk was not available. Refer to the following footnote with regard to Wellington Free and the 5% of theoretical eligible calls. The remaining 4% difference is the gray calls that were not eligible for secondary triage in the Wellington Free - Medibank pilot. 45 These calls were not transferred to secondary triage during the pilot mainly due to different training systems between the three Communication centres, call takers forgetting to route calls to Healthline, and call takers making judgments based on their experience of calls returned from Healthline. On rare occasions, Healthline was not available or the caller refused to be put through to Healthline. 46 It is important to note that this statement applies at the time of the call. It is unknown whether, and if so, how many of the callers subsequently recontacted 111. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 27

31 Chart 3: Resolution of secondary triage calls 70% 54% 46% 30% Resolved by hear and treat (did not require an ambulance) St John (n=1,641) Returned to 111 (usually resulting in dispatch of an ambulance) Medibank (n=2,713) The differing resolution rates reflect the different approaches of the two pilots to secondary triage as described previously in KEQ1: Clinical safety, different data collation and a different range of dispositions. For example, alarm calls are referred to secondary triage in Christchurch and not in Wellington. St John reports on three dispositions immediate GP (within 3-4 hours), delayed referral and self-care. Healthline reports in detail on a wider range of dispositions and the various timeframes recommended (e.g. see GP immediately, within 4 hours, within 8 hours, within 24 hours, within 72 hours, within 2 weeks). The following chart illustrates the level of care to which people were triaged, for those resolved by secondary triage. There were marked differences between the two pilots reflecting their different approaches and data recording systems. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 28

32 Chart 4: Level of care for those resolved by secondary triage 47, 48 47% 34% 29% 37% 18% 8% 12% 16% 0% 0% Attend ED immediately Call or see GP (or other health provider) immediately or within 4 hrs See GP (or other health provider) within 24 hrs See GP (or other health provider) within 2 weeks Home / self-care St John (n=492) Medibank (n=1,141) Impact on ambulance dispatch 49, 50 The impact of the number of calls resolved via secondary triage meant that St John reduced the overall dispatch of ambulances by four percent and Wellington Free Ambulance by just over six percent, for their respective pilot areas 51. This represents approximately one year s growth in demand based on an average growth rate of 4.5% 52. This is a practically significant result, creating efficiencies that will allow growth to be managed. 47 As already noted, the two pilots had different dispositions and used different reporting categories. This chart combines a number of categories in order to provide a comparative picture. The category call or see GP (or other health provider) immediately or within 4 hrs includes Mediban s categories for see GP or other health provider immediately and within 4 hours. 48 The Medibank total of 1,141 excludes 266 cases that were counted as resolved by secondary triage by Wellington Free Ambulance (their total was 1,473 based on the number of calls that did not return from Healthline). The 266 cases include 144 not recorded, 83 who declined triage, 27 for whom 111 was activated and 12 cases where the caller hung up. 49 The evaluative rubrics included the criterion practically significant reductions are achieved in the ambulance transport rate for the area. Data was provided by the two pilots which enabled calculation of the transport rate as a (i) percentage of total calls and (ii) percentage of ambulance dispatches (St John) or ambulance response (Wellington Free Ambulance), and a pre-pilot comparison with the same period a year earlier. The specific impact of secondary triage was unable to be determined as the transport rate is affected by incidents that are resolved by ambulance crew at the scene ( see and treat ), the increasing use of first responder cars who resolve incidents via see and treat, and when a dispatched ambulance is no longer required. The providers noted they expected that the ambulance transport rate calculated as a percentage of ambulance dispatches would increase as a result of secondary triage as dispatches/responses are to incidents more likely to require ambulance transport to the emergency department. 50 The evaluative rubrics also included the criterion ambulance resources from low acuity work to the extent there is a significant improvement in the availability of ambulances and crews for more timely dispatch to higher acuity cases. Due to the reprioritisation of the priority codes that was implemented in August 2012, the evaluation was unable to clearly determine whether this occurred. Over the country (and in the St John pilot area), the time taken for an ambulance dispatch to purple and red codes got shorter, and orange, green and grey have got longer, as was expected. 51 The two providers calculate the total number of calls differently so these figures are indicative. 52 Annual growth figure supplied by NASO. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 29

33 Table 2: Reduction in avoidable dispatch of ambulances St John pilot Wellington Free pilot Total number of calls 12,317 23,188 Number resolved by 492 1,473 secondary triage % ambulance not dispatched 4% 6.3% In terms of the impact on ambulance dispatch by acuity, both St John and Wellington Free Ambulance data show a significant reduction in the grey code. Table 3: St John pilot: Change in ambulance dispatches by acuity 53 Priority codes Pre-pilot (Oct 11 Feb 12) Pilot (Oct 12 Feb 13) Change in percent points Purple % % + 9% Red 4,694 98% 4,668 97% - 1% Orange 2,510 94% 4,232 94% 0 Green 1,485 96% 1,916 94% - 2% Grey 2,632 95% % -15% Note the Grey calls in the pre-pilot period include ALL grey calls, and in the pilot period, only those calls coded Grey which were not triaged. Table 4: Wellington Free Ambulance: Change in ambulance response by acuity Priority codes Pre-pilot (Oct 11 Feb 12) Pilot (Oct 12 Feb 13) Change Purple % % - 2% Red 7,897 95% 8,158 94% - 1% Orange 4,796 93% 5,828 91% - 2% Green % 2,507 85% - 8% Grey 1,845 96% % - 7% Secondary Triage Grey 1,672 90% 1,120 29% - 61% Impact on emergency departments NASO and the Ministry of Health were also interested in whether there was any impact on emergency department (ED) resources. Specifically whether there was a reduction in the avoidable presentation to ED via ambulance. This was unable to be substantively answered from the available data. 53 In tables 3 and 4, the colour-based system of priority codes did not exist, so the allocation is a best guestimate. The n in the tables is the number of ambulance dispatches (St John) or the number of ambulance responses (Wellington Free Ambulances). The percentages are calculated as a proportion of the total number of calls (which is not shown) allocated within each priority code. For example, in Table 3, St John dispatched 130 ambulances to purple incidents during the pre-pilot period which was 90% of the total number of purple coded calls (130/144 x 100 = 90%). Evaluation of the secondary triage pilots in the Ambulance Communications Centres 30

34 Determining the impact of secondary triage on ED resources would require knowing how many people were transported by ambulance to ED prior to secondary triage, and how many are now triaged to different levels of care due to secondary triage. It would also require knowing how many people are advised (as part of secondary triage) to attend ED immediately via their own resources (one of the Healthline secondary triage dispositions), and how many others present at ED who were advised to do something else. A potential proxy of the impact on ED resources could be the reduction in ambulance dispatches, with the Wellington Free Ambulance-Healthline figures adjusted to include the disposition attend ED immediately via own resources, as follows. Table 5: Indicative reduction in avoidable presentation to ED 54 Christchurch pilot Wellington pilot Total number of calls 12,317 23,188 Number resolved by 492 1,267 secondary triage (without ambulance dispatch) or advised to attend ED via own resources % proxy of reduced presentation to ED 4% 5.5% However, this proxy needs to be treated with the utmost caution and would require further investigation before being used in any substantive sense 55. KEQ 3: What are the efficiency implications of the pilot processes and outcomes? Savings from the secondary triage pilots include a decrease in ambulance dispatches that represents approximately one year s growth in demand, estimated savings of $109,000 on a marginal cost basis or $1.375 million on an average cost basis (over the 5 month pilot period), and improved health system efficiencies with the potential for further improvement. 54 This has been calculated by adjusting the Wellington Free Ambulance figures to include the 206 callers that were referred to Healthline and advised to attend their emergency department immediately. One of the differences between the two pilots is the disposition attend emergency department (ED) immediately, which means the caller doing this via their own transport resources. St John dispatches an ambulance for callers who need to attend ED immediately, where as Healthline separates callers needing to attend ED into those that require an ambulance dispatch (returned to 111) and those that can attend via their own resources. 55 Reviewers of the report were divided as to whether this proxy had legitimacy. While all acknowledged this made sense in principle (i.e. a reduction in ambulance dispatch and subsequent transportation of patients would be reflected in reduced ED attendance figures), they noted this is notoriously difficult to prove. An example was provided where the Safer Cities triage centre in Courtenay Place - an initiative designed to reduce the impact on ED attendance, did not result in ED noticing any difference (Gary Strong). Views ranged from yes it does have credibility, in the absence of anything else, this is the best that can be done to removing the proxy which looks to demonstrate avoidable ED attendance. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 31

35 The following issues were explored as part of the evaluation: Operational costs of call triaging Impacts on wider health system costs Implications for health system efficiency. Operational costs of call triaging This section addresses two questions, the first being the operational costs of the two pilots compared to the status quo. For both St John and WFA, the operational costs of the pilots were covered from within existing budgets. The operational cost of the secondary triage calls to Medibank required additional funding from the Ministry of Health. The second question explored was the operational cost implications of taking the pilot secondary triage models to scale. The providers identified that taking secondary triage to scale would involve the following costs: One-off implementation costs such as: Staff training (including a more comprehensive national training package) Public relations campaign (to manage public expectations of the new system) Alignment of secondary triage and ambulance communication centres processes and language (for example, priority code for those returned to 111 from secondary triage) Integration of secondary triage and ambulance data systems (to enable the collection and analysis of a common set of data) Possible additional cost if a new or modified secondary triage tool is developed Ongoing costs 56 An increase in call taking staff and managers (for both Medibank and the ambulance communication centres) An increase in average call time for Healthline nurses (due to the comprehensive handover procedures between 111 and Healthline). With regard to the longer-term goal of an integrated health system, integrating the information technology systems of the ambulance sector and secondary triage providers would potentially be a significant future cost, if this is possible Medibank has undertaken modeling of the anticipated additional call volume and costs of providing secondary triage to St John s Northern Ambulance Communications Centre. With regard to commercial sensitivities, it is suggested that the Ministry of Health approach Medibank directly for this information if required. 57 Medibank and Wellington Free Ambulance have discussed the potential of doing this in the future. Medibank have trialed integrating the two systems in Australia but found it was too difficult and costly. This has yet to be successfully achieved anywhere in the world. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 32

36 Impacts on wider health system costs This section was concerned with any impacts the secondary triage models may have had on selected factors 58 that influence wider health system costs, compared to the status quo. The secondary triage pilots resulted in a reduction of between 4-6% in ambulance dispatch between the two pilots. This represents approximately one year s growth in demand based on an average growth rate of 4.5% 59. The estimated savings on a marginal cost (short-run) basis is approximately $109,000. The estimated savings on an average cost basis (which takes into account the eventual need to add an extra crewed ambulance in response to growth) is approximately $1.375 million. On an annualized basis, this translates to $262,000 based on marginal cost or $3.3 million based on average cost. Table 6: Estimated marginal and average cost savings over the 5 month pilot period 60 Total number resolved by Marginal cost Average cost secondary triage St John = 492 $55.60 $700 WFA = 1,473 x 1,965 x 1,965 Total = 1,965 $109,254 $1,375,500 Implications for health system efficiency The findings from the evaluation show that secondary triage can improve health system efficiencies, with the potential for further efficiencies to be realised. Impact on people People are being directed to a lower, safe level of care, and there is potential for this to occur to a greater extent 61. Secondary triage was found to be clinically safe. Overall, secondary triage shortened people s journey to the most appropriate level of care. There may be a differential social impact given that secondary triage shifts costs from the public health system to individuals, for example, people being required to 58 The available data meant that definitive conclusions were not possible about the impact of secondary triage on factors such as ambulance transports and the number of low acuity patients taken to emergency departments by ambulance, as originally anticipated in the evaluative rubrics. In terms of the impact on ambulance dispatch by acuity, both St John and Wellington Free Ambulance data show a significant reduction in the grey code (refer earlier section KEQ 2: Impact on ambulance dispatch). 59 Annual growth figure supplied by NASO. 60 The marginal and average cost figures were supplied by NASO in consultation with the two ambulance providers. The marginal cost figure is based on calculations undertaken by St John and WFA, and the average cost figure is anecdotal and indicative of the general order of magnitude. 61 As the providers determine the codes best suited to secondary triage (with regard to the triage approach offered by Medibank), and given there was a high level of default to a see and treat approach in the St John pilot. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 33

37 use their own resources to travel to and pay for GP and other health provider services. Some concern was identified about secondary triage affecting access to health care for people with less financial and transport resources. Savings and more efficient use of services There are practically significant savings from the reduction in ambulance dispatches. The savings represent approximately one year s growth in the demand for ambulance services. There is the potential for further savings to occur (refer to footnote 33). It also appears that ambulance services are being freed up from low acuity work, which in turn will enable a faster response to higher acuity cases. KEQ 4: What can be learned from the evaluation to inform future decisions? Success of the pilots has been demonstrated by the realisation of two key objectives: the lowest safe level of care appropriate to clinical need for users, and health system efficiencies. MoH was interested in whether the outcomes were better/worse/the same as the status quo (see pp 8, 9). The pilots have delivered outcomes better than the status quo, as demonstrated by: the clinical objective of the lowest safe level of care appropriate to clinical need being met, and the achievement of health system efficiencies, particularly in relation to the 4-6% reduction in the use of ambulances (which will manage growth in the sector). The pilots have provided proof of concept in relation to secondary triage. A further point of interest for MoH was any service provision, service user safety, or cost implications of expanding either pilot option across New Zealand. This section considers the learnings from the pilots that will assist in identifying the policy and funding work that will be required if secondary triage is to be rolled out. The following issues are explored as part of this discussion: Success factors Considerations for the future of secondary triage. Success factors The focus in this section is on what can be learned about why and how the models are effective in each location. What worked well The success of the pilots in both locations has been enabled by organisational commitment by all three providers, and committed and capable key individuals responsible for ensuring the effective operation of the pilots. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 34

38 In both locations there has been seamless transfer of calls between the communications centre and secondary triage. In Wellington seamless transfer has been facilitated by the warm transfer process and by a direct line back to CentralComms from Medibank. In Christchurch seamless transfer is facilitated by the CAD to CAD system which means that the data gathered by the 111 call-taker is available to the paramedic so reduces repetition of questions, and the location of the secondary triage paramedic within the call centre enables contact with the calltaker if appropriate. The CECC triage tool used by Medibank is considered to be a strength of the Wellington model. The tool is externally validated, is used internationally, and has robust updating and localising processes. Users in the phone survey commented on the efficiency and competence of the Medibank nurses. A strength of secondary triage in Christchurch has been a people-centred focus on the patient receiving and accessing appropriate care. This includes the availability of a range of dispositions at the local level, and the facilitation of transport options so that users can access the recommended actions. Users in the phone survey described one of the best things about the process was that the paramedics were reassuring. Integral to the success of the pilots were the robust clinical governance and risk management frameworks and processes which the three providers applied to ensure the safety of secondary triage for 111 callers. Learnings The evaluation highlighted different models in relation to the boundaries of secondary triage. The core concept of secondary triage for the pilots was well defined - low acuity calls were transferred to a nurse or paramedic on the phone to determine the course of action most appropriate to the person s clinical need. In the St John model the course of action deemed appropriate in many instances was to dispatch a single paramedic for a see and treat assessment (which they considered more appropriate when dealing with more complex determinants). There are two pertinent comments here. First, as previously noted, it has meant that in the pilot phase, St John has not really tested a discrete secondary triage process. On the other hand, St John has demonstrated the value of having a greater range of options at their disposal (including see and treat options for assessment that can be organised rapidly), when a simple hear and treat solution is deemed insufficient. Things that did not work so well, as discussed throughout the report include: The interface between ProQA and secondary triage contributing to confusion for some callers, as a result of required ProQA scripting (currently being addressed) Evaluation of the secondary triage pilots in the Ambulance Communications Centres 35

39 Having secondary triage half time (in Christchurch), competing with the clinical desk paramedic s priority to support paramedics on the road Possible gaps in the CAST tool itself, and ad hoc application of the CAST tool by secondary triage paramedics A level of user dissatisfaction with the secondary triage process, perhaps due to a lack of public information about secondary triage and therefore understanding of the reasons and advantages of being transferred to secondary triage. There was also an indication that secondary triage worked best for those 111 calls that were less serious/urgent The level of training and support provided to paramedics to work in a telephone triage environment and in the application of the CAST tool. For Medibank nurses it was suggested that a familiarization with ambulance codes, and possibly additional training in triaging emergency situations could be useful Some gatekeeping of access to transport and attitude issues by Medibank nurses. Key contributors to effectiveness Key features of the pilots that appear to be important contributors to their effectiveness emerge from the above findings. These include: The importance of a key manager in each provider organization who understands the business well, and secondary triage, so is able to hold onto the big picture while attending to operational issues The need for users to understand what secondary triage is, its purpose and why they are being transferred The value of a systematic approach to secondary triage, supported by good IT-based decision support A seamless transfer of calls Training, support and ongoing staff development in telephone triage generally, and emergency medicine specifically A responsive secondary triage person who can balance safe outcomes for the user/patient (including enabling patients/users to access the recommended course of action) with operational imperatives, such as timeliness A 24/7 service and a quick response to secondary triage calls. Economies of scale at Medibank mean that nurses are able to respond to secondary triage calls quickly Robust clinical governance and risk management processes. Considerations for the future of secondary triage The issues that are covered in this section have emerged from the analysis of the data, discussions with stakeholders in the course of the evaluation, and discussions with stakeholders at the topline findings workshop. This section focuses on features of the model, in addition to those identified above, that may need to be changed (e.g. dropped, modified or created) to enable scaling up and implementation in other regions. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 36

40 All stakeholders consulted are in agreement that a single secondary triage approach across the country would be ideal for secondary triage if scaled up, in the same way as the ambulance communications centres operate with a single approach. Some of the problems for call-takers, associated with the implementation of secondary triage, have arisen because of the different steps that they need to follow, dependent on where the 111 caller is located, to deal with ProQA determinants coded grey or gray. For a single approach to work well providers would need to reach agreement on: a single tool classification of determinants that can be effectively dealt with by hear and treat the transfer process used between 111 and secondary triage (warm transfer as used in Wellington or call-back as used in Christchurch). Wellington providers and MoH commented that a warm transfer is good from both a clinical safety perspective (as the call is less likely to be lost) and a quality perspective (as a calltaker will know that the transfer of the call has taken place). St John reports that their medical advisor states that both approaches are safe. No research was known that compared the impact of each approach on call-taking, and a robust comparison of these approaches might be useful in future. whether there is a preference about where secondary triage clinicians should be located. Questions have been raised about whether nurses could be located in the Ambulance Communications Centres, and similarly, if Medibank might employ paramedics. Both of these issues require further consideration. There was also agreement that roll-out of secondary triage needs an infrastructure in place to support it. This would include: o a range of local health services. Secondary triage appeared to work best where there was a range of local health services to which users/patients could be referred. o a range of transport options to enable users to access recommended actions from secondary triage. There was wide agreement amongst providers and MoH that access to a wide range of transport options was desirable to support people to access the recommended action following secondary triage. Currently this range of options isn t available and would need to be developed for an effective roll-out of secondary triage. The evaluation highlighted different approaches to secondary triage, on a continuum from a more process-oriented approach to a more diagnostic approach. At the topline findings workshop there was consideration of the possibility that both approaches may be useful in particular circumstances. Feedback from the small group of service users surveyed indicated that satisfaction with a process-oriented approach was higher for cases that users considered to be of a less serious or urgent nature. This would suggest that a diagnostic approach is better for more serious cases, but these ideas would need to be tested further. Investigation would also be needed to identify the ProQA determinants which may be better suited to resolution by hear and treat and are relatively easy to close off, and those which may require a more diagnostic approach. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 37

41 The ways in which providers currently capture data about the use of ambulances and secondary triage are not easily comparable. For any roll-out of secondary triage there would need to be some standardization of data capture and analysis across providers. As noted earlier in the report, there also need to be better processes for monitoring the health outcomes of people using secondary triage. While this may be facilitated in the longer term when/if NHI data is able to be shared across services, in the short to medium term there is a need to capture data on the number of re-presentations, where a second 111 call has been received after secondary triage. The biggest thing that needs to change if secondary triage is to be rolled out is user/patient perception of secondary triage. At the time the pilots were introduced, there was no information made available to the public about this change to the ambulance sector. Users who were surveyed reported feeling confused about what was happening. Amongst stakeholders there is agreement that a public awareness campaign about secondary triage will be important for any scaling up of the pilots. Evaluation of the secondary triage pilots in the Ambulance Communications Centres 38

42 Appendices Appendix One: Process maps of secondary triage in practice Evaluation of the secondary triage pilots in the Ambulance Communications Centres 39

43 Evaluation of the secondary triage pilots in the Ambulance Communications Centres 40