Page 1 of 23
Quality Ambition: Safe NHS Lanarkshire aims to be the safest health and care system in Scotland with no avoidable deaths, reduction in avoidable harm, a sustainable infrastructure for patient safety and improvement, a sustainable safety culture, and safe, reliable and person centred care. Safe: Percent Unadjusted Inpatient Mortality (NHSL Acute Hospitals) Percent Unadjusted Inpatient Mortality (NHSL Acute Hospitals) Total number of in-hospital deaths Total number of deaths (TD) + live discharges (LD) NHSL unadjusted inpatient mortality has decreased in June in comparison to the previous month. No shift or trend in the data is noted at present. With effect from May 2015, Mental Health deaths and admissions in the three acute sites are now included in this measure. This is as a result of these locations now being recorded via TrakCare Patient Management System. Safe: Percent Unadjusted Inpatient Mortality (Hairmyres) Percent Unadjusted Inpatient Mortality (Hairmyres) Total number of in-hospital deaths Total number of deaths (TD) + live discharges (LD) Hairmyres unadjusted inpatient mortality has decreased in June to below the median line, ending the previously noted upward shift in the data. January saw the highest reported % in the two year reporting period (3.52%). Review of case listings for this period identified a high % of frail elderly patients and patients with cancer diagnosis, many of whom had died after 30 days in hospital and were part of the high numbers of delayed discharges on the site that are now much reduced. With effect from May 2015, Mental Health deaths and admissions in the three acute sites are now included in this measure. This is as a result of these locations now being recorded via TrakCare Patient Management System. UIM_NHSL Unadjusted Inpatient Mortality - NHSL UIM_HM Unadjusted Inpatient Mortality - Hairmyres 4.0 4.0 3.5 3.5 3.0 3.0 2.5 2.5 % 2.0 % 2.0 1.5 1.5 1.0 0.5 0.0 % Median Jun-15 1.0 0.5 0.0 % Median Jun-15 Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median is calculated using 2008 unadjusted mortality. Description: Unadjusted mortality is a complementary measure to HSMR. This provides an indicator of trend. The trend should be decreasing and under the median line. Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median is calculated using 2008 unadjusted mortality. Description: Unadjusted mortality is a complementary measure to HSMR. This provides an indicator of trend. The trend should be decreasing and under the median line. Page 2 of 23
Safe: Percent Unadjusted Inpatient Mortality (Monklands) Percent Unadjusted Inpatient Mortality (Monklands) Total number of in-hospital deaths Total number of deaths (TD) + live discharges (LD) Monklands unadjusted inpatient mortality has increased slightly in June in comparison to the previous month. The % dropped in March 2015 ending the previously noted upward trend in the data. No shift or trend in the data is noted at present. With effect from May 2015, Mental Health deaths and admissions in the three acute sites are now included in this measure. This is as a result of these locations now being recorded via TrakCare Patient Management System. Safe: Percent Unadjusted Inpatient Mortality (Wishaw) Percent Unadjusted Inpatient Mortality (Wishaw) Total number of in-hospital deaths Total number of deaths (TD) + live discharges (LD) Wishaw unadjusted inpatient mortality has increased slightly in June in comparison to the previous month. The percentage increased in December to above the median line ending the downward shift in the data noted in the previous 9 months of data (six or more consecutive points below the median line). No shift or trend in the data is noted at present. With effect from May 2015, Mental Health deaths and admissions in the three acute sites are now included in this measure. This is as a result of these locations now being recorded via TrakCare Patient Management System. UIM_MK Unadjusted Inpatient Mortality - Monklands UIM_WG Unadjusted Inpatient Mortality - Wishaw 4.0 4.0 3.5 3.5 3.0 3.0 2.5 2.5 % 2.0 % 2.0 1.5 1.5 1.0 0.5 0.0 % Median Jun-15 1.0 0.5 0.0 % Median Jun-15 Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median is calculated using 2008 unadjusted mortality. Description: Unadjusted mortality is a complementary measure to HSMR. This provides an indicator of trend. The trend should be decreasing and under the median line. Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median is calculated using 2008 unadjusted mortality. Description: Unadjusted mortality is a complementary measure to HSMR. This provides an indicator of trend. The trend should be decreasing and under the median line. Page 3 of 23
Safe: Hospital Standardised Mortality Ratio (HSMR) - Hairmyres Number of Observed Deaths versus Predicted Deaths (30 days) -Hairmyres Number of Observed deaths and Number of Predicted deaths Not Applicable The number of observed deaths has increased at Hairmyres in comparison to the previous quarter and is parallel with predicted deaths which have also increased. This is reflected in the HSMR figure for this quarter. Hospital Standardised Mortality Ratio - Hairmyres Number of Observed deaths Number of Predicted deaths The HSMR for Hairmyres has remained the same as the previous quarter (0.82). The last 8 data points (quarterly) have remained below the median (and below 1.0) which is indicative of a downward shift in the data (six or more consecutive points below the median line). PREDD_HM OBSD_HM Observed Deaths versus Predicted Deaths - Hairmyres HSMRM_HM HSMR - Hairmyres 320 1.2 280 1.0 240 Number 200 160 120 Rate 0.8 0.6 0.4 80 40 0.2 0 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 0.0 Jul-09 Oct-09 Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 Observed Deaths Predicted Deaths HSMR Median Description: Predicted probability of death is calculated based on primary diagnosis; specialty; age; sex; where admitted from; number and severity of prior morbidities; number of emergency admissions in previous 12 months; whether admitted as an inpatient or day case and type of admission (elective / non-elective). Data reliability = Very High (see appendix c) Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The median is calculated in line with the ISD HSMR Dashboard. Description: HSMR is calculated as: Observed Deaths within 30 Days / Predicted Deaths Within 30 Days and is calculated for all acute inpatient and day case patients admitted to all specialties. Data reliability = Very High (see appendix c) Page 4 of 23
Safe: Hospital Standardised Mortality Ratio (HSMR) - Monklands Number of Observed Deaths versus Predicted Deaths (30 days) -Monklands Number of Observed deaths and Number of Predicted deaths Not Applicable The number of both observed and predicted deaths has decreased at Monklands in comparison to the previous quarter with a greater reduction in observed deaths than predicted. This is reflected in the HSMR figure for this quarter. Hospital Standardised Mortality Ratio - Monklands Number of Observed deaths Number of Predicted deaths Monklands HSMR has decreased in Oct-Dec 2014 to 0.91 and this is the 4th consecutive quarter where the HSMR for Monklands has been less than 1.0 PREDD_MK OBSD_MK Observed Deaths versus Predicted Deaths - Monklands HSMRM_MK HSMR - Monklands 320 1.2 280 240 1.0 Number 200 160 120 80 Rate 0.8 0.6 0.4 40 0.2 0 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 0.0 Jul-09 Oct-09 Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 Observed Deaths Predicted Deaths HSMR Median Description: Predicted probability of death is calculated based on primary diagnosis; specialty; age; sex; where admitted from; number and severity of prior morbidities; number of emergency admissions in previous 12 months; whether admitted as an inpatient or day case and type of admission (elective / non-elective). Data reliability = Very High (see appendix c) Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The median is calculated in line with the ISD HSMR Dashboard. Description: HSMR is calculated as: Observed Deaths within 30 Days / Predicted Deaths Within 30 Days and is calculated for all acute inpatient and day case patients admitted to all specialties. Data reliability = Very High (see appendix c) Page 5 of 23
Safe: Hospital Standardised Mortality Ratio (HSMR) - Wishaw Number of Observed Deaths versus Predicted Deaths (30 days) - Wishaw Number of Observed deaths and Number of Predicted deaths Not Applicable The number of both observed and predicted deaths has increased at Wishaw in comparison to the previous quarter, with more observed deaths than predicted. This is reflected in the HSMR figure for this quarter. Hospital Standardised Mortality Ratio - Wishaw Number of Observed deaths Number of Predicted deaths There has been an increase in the Oct-Dec 2014 HSMR (1.06), taking the rate above the median and ending the previous positive shift. A casenote review has been completed to establish any learning. This data point is the subject of increased scrutiny and a separate report has been considered by HQAIC. In addition, a review of surgical deaths that quarter is due to take place.the Board is working with colleagues from HIS and Public Health Intelligence to further review the data to better understand HSMR related analyses including the relationship with unadjusted mortality. PREDD_WG OBSD_WG Observed Deaths versus Predicted Deaths - Wishaw HSMR_WG HSMR - Wishaw 320 1.2 280 1.0 240 Number 200 160 120 Rate 0.8 0.6 0.4 80 40 0.2 0 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 0.0 Jul-09 Oct-09 Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 Observed Deaths Predicted Deaths HSMR Median Description: Predicted probability of death is calculated based on primary diagnosis; specialty; age; sex; where admitted from; number and severity of prior morbidities; number of emergency admissions in previous 12 months; whether admitted as an inpatient or day case and type of admission (elective / non-elective). Data reliability = Very High (see appendix c) Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The median is calculated in line with the ISD HSMR Dashboard. Description: HSMR is calculated as: Observed Deaths within 30 Days / Predicted Deaths Within 30 Days and is calculated for all acute inpatient and day case patients admitted to all specialties. Data reliability = Very High (see appendix c) Page 6 of 23
Safe: Coding Completeness - NHSL Discharges coded within 6 weeks in NHSL Lanarkshire Total number submitted within 6 week target Total number of discharges expected to be submitted The technical issue that prevented SMR data to be extracted on time for the February target date was resolved in March allowing NHSL to submit on time. For May 2015 SMR01 records submitted by the target date of 6 weeks following the end of the month of discharge has remained at 88%, same figure as the previous month. The data demonstrates 3 data points above the median (March - 92%, April - 88%, May - 88%, median is 85%). Safe: Coding Completeness - Per Site Discharges coded within 6 weeks in NHSL Lanarkshire Total number submitted within 6 week target Total number of discharges expected to be submitted In May, both Hairmyres and Wishaw have decreased slightly since last month, by 1%; Monklands however has shown an increase of 2% since last month.work is continuing at all three sites with a focus on: sessions at Wishaw to raise awareness and provide education on discharges and coding, further development and testing of new MiLAN monitoring report and use of the aide memoire (TESTS) to support the administration and secretarial staff when producing discharge summaries. Further work to support the improvement of the quality of discharges for deceased patients will be to use the new death certification process; roll out of the mortality and morbidity review form and named lead consultant for each ward. Process Measure Process Measure DISCH_NHSL % Discharges coded within 6 weeks in NHS Lanarkshire DISCH_NHSL % Discharges coded within 6 weeks - All Hospitals in NHS Lanarkshire 100 100 80 80 60 60 % % 40 40 20 0 Jun-13 % Median Goal Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median was calculated using Jan-Dec 2012 data. Following a shift the median has been recalculated from Dec 2013. Description: The Scottish Government target for SMR submission to ISD is 6 weeks (42 days) following discharge/transfer/death. ISD calculates timeliness as data received 6 weeks following the end of month of discharge/transfer/death e.g:. all SMR01 records with a March 2014 date of discharge/transfer/death would be expected to be submitted to ISD by 12th May 2014. The data on the chart above demonstrates the number of SMR01 records submitted by the target date of 6 weeks following the end of the month of discharge. Data reliability = High (see appendix c) 20 0 Jun-13 % Hairmyres % Monklands % Wishaw Goal Description: The Scottish Government target for SMR submission to ISD is 6 weeks (42 days) following discharge/transfer/death. ISD calculates timeliness as data received 6 weeks following the end of month of discharge/transfer/death e.g:. all SMR01 records with a March 2014 date of discharge/transfer/death would be expected to be submitted to ISD by 12th May 2014. The data on the chart above demonstrates the number of SMR01 records submitted by the target date of 6 weeks following the end of the month of discharge. Data reliability = High (see appendix c) Page 7 of 23
Safe: Rate of Harms (NHSL) Rate of Harms (NHSL) per 1000 total deaths and live discharges Total number of Falls resulting in injury from DATIX, Pressure Ulcers reported on DATIX and Cardiac Arrests (CA audit data) Total number of deaths & live discharges for the same period The rate of harms for April 2015 has decreased in comparison to the previous month. The chart is based on 3 of the 4 harms that the reducing harm collaborative teams are focusing on as it has not yet been possible to get organisation level data on CAUTI. Work is underway to get this data in the required format to incorporate it into this measure. ROH_NHSL 14 Rate of Harms (NHSL) 12 10 Rate 8 6 4 2 0 May-13 Jun-13 Rate Median Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median is calculated using Aug 12-Jul 13 data points. Description: As part of the Boards Prioritised patient safety plan measuring the number of harms and achieving 95 % harm free care are both strategic aims locally. These measures are key drivers of the National Scottish Patient Safety Programme Phase 2 Page 8 of 23
Safe: CAUTI Bundle Reliability % Reliable implementation of the catheter insertion and maintenance bundles Total number of patients sampled where catheter insertion and maintenance bundle was reliably implemented Weekly sample of all patients with a catheter in situ or removed within previous 48hrs (5 patients per ward per week) The Microsystems data represented is for one of the pilot teams at Wishaw General. As part of the Reducing Harm Collaborative, pilot teams are working towards achieving the aim of a 30% reduction in CAUTI and are working on testing and reliably implementing the catheter insertion and maintenance bundles. These charts demonstrate the progress this team is making in improving the reliability of key processes known to impact on the outcomes. Process Measure Safe: Number of catheters / CAUTI's in Ward / Caseload per week Total number of catheters / CAUTI's in ward / caseload per week Total number of catheters / CAUTI's in ward / caseload per week Not applicable The Microsystems data represented is for one of the pilot teams at Wishaw General. Teams are measuring weekly numbers of catheters and CAUTI. The rate of CAUTI remains consistently low but there is an opportunity to reduce the number of patients with a catheter which in turn will reduce CAUTI. The reduced number of catheters reported in this team may be due to increased awareness of the issue, with staff testing ways to reduce unnecessary catheterisation and remove catheters timeously. Page 9 of 23
Safe: Falls Bundle Reliability % Reliable implementation of the falls bundle Total number of patients sampled where falls bundle was reliably implemented Weekly sample of all patients in ward The data represented is for the one of the pilot teams at Monklands. As part of the Reducing Harm Collaborative, pilot teams are working towards achieving a 20% reduction in falls with harm and 25% reduction in falls and are working to test and reliably implement the Falls Bundle. This team have been testing improvements in each element of the falls bundle processes. The focus recently has been on improving Lying and Standing BP reliability using a separate form/coloured paper. Their falls team has changed and the SCN indicated this is the reason behind the variation previously. This has improved recently with a new CSW and Staff nurse involved. The team are raising awareness of the work among the wider team hence the early improvement in process reliability data. Process Measure Safe: Number of Falls in Ward / Caseload per week Total number of falls in ward / caseload per week & Total number of falls with harm in ward / caseload per week Total number of falls in ward / caseload per week & Total number of falls with harm in ward / caseload per week Not applicable The Microsystems data represented is for one of the pilot teams at Monklands Hospital. Teams are measuring weekly numbers of falls and falls with harm alongside reliability of key processes. The Patient Safety Team are working with teams to increase engagement and ownership of the safety improvement work and ensure all teams are actively participating, measuring and testing changes. Page 10 of 23
Safe: Pressure Ulcer Bundle Reliability % Reliable implementation of the Pressure Ulcer bundle Total number of patients sampled where the Pressure Ulcer Bundle was reliably implemented Weekly sample of all patients at risk of pressure ulcer (5 patients per ward per week) The Microsystems data represented is for one of the pilot teams at Hairmyres Hospital. As part of the Reducing Harm Collaborative, pilot teams are working towards achieving the aim of zero pressure ulcers or 300 days between hospital acquired pressure ulcers per ward and are working on testing and reliably implementing the SSKIN bundle. These charts demonstrate the progress this team is making and early signs of improvement in key processes known to impact on the outcomes. The teams are being encouraged to focus on testing changes to further increase the reliability in processes. Process Measure Safe: Number of Pressure Ulcers in Ward / Caseload per week Total number of Pressure Ulcers in ward / caseload per week Total number of Pressure Ulcers in ward / caseload per week Not applicable The Microsystems data represented is for the pilot ward at Hairmyres Hospital. Teams are measuring weekly numbers of Pressure Ulcers either inherited to the ward or acquired in the ward. Page 11 of 23
Safe: Cardiac Arrest Rate - NHSL Cardiac Arrest Rate (True cardiac +/- respiratory arrests) per 1000 total deaths and live discharges (NHSL) Total number of true cardiac +/- respiratory arrests Total number of deaths & live discharges for the same period To date there is no statistically significant improvement or reduction in Cardiac Arrests evident. Pilot teams are testing and implementing changes to increase the reliability of key processes that are recognised to impact on and achieve a reduction in Cardiac Arrests. Learning from a previous cardiac arrest casenote review highlighted the following areas for improvement: Reliability of observations, robust DNACPR, End of Life Care and Ceiling of Treatment processes and reliable recognition and response.the reporting of MEWS whilst an important and ongoing component of the work; has been replaced to represent the current focus of improvement work which is ensuring reliability of clinical observations which in turn will increase the reliability of MEWS as part of the recognition and response processes. Aligning the work of DNACPR and associated measures has been identified as an additional area of benefit in progressing this agenda. Safe: Cardiac Arrest Rate - Hairmyres Cardiac Arrest Rate (True cardiac +/- respiratory arrests) per 1000 total deaths and live discharges (Hairmyres) Total number of true cardiac +/- respiratory arrests Total number of deaths & live discharges for the same period This chart show normal variation. Page 12 of 23
Safe: Cardiac Arrest Rate - Monklands Cardiac Arrest Rate (True cardiac +/- respiratory arrests) per 1000 total deaths and live discharges (Monklands) Total number of true cardiac +/- respiratory arrests Total number of deaths & live discharges for the same period Safe: Cardiac Arrest Rate - Wishaw Cardiac Arrest Rate (True cardiac +/- respiratory arrests) per 1000 total deaths and live discharges (Wishaw) Total number of true cardiac +/- respiratory arrests Total number of deaths & live discharges for the same period This chart show normal variation. This chart currently displays normal variation. Attention is drawn to the Jan 2015 data point that has breached the upper control limit due to the late inclusion of two cardiac arrests occurring during that month. A review of the data collection process is underway to establish potential delays in the process and areas for improvement. A cardiac arrest case note review is also scheduled to take place in August to establish any learning from the higher number of arrests that month. Findings should be available 1st week of September. Page 13 of 23
Safe: Number of Patient Safety Leadership Walkrounds - NHSL Number of Patient Safety Leadership Walkrounds (NHSL) The number of Patient Safety Leadership Walkrounds taking place Not applicable Patient safety walkrounds are recognised as a useful activity that demonstrates an organisations commitment to making patient safety a priority. Through the process safety issues raised are turned into key actions with associated timescales to ensure these are progressed in a timely and appropriate manner. NHS Lanarkshire aim to complete 1 patient safety leadership walkround per week and this is being achieved. Safe: Patient Safety Leadership Walkrounds - Action Status - NHSL Patient Safety Leadership Walkrounds - Action Status (NHSL) Total Number of Actions Total Number of Actions Closed / Closed with monitoring / Open (overdue) / Open (within target) There has been a significant increase in the closing of actions since the start of the reporting period and as a result the organisation has a deeper understanding of overdue actions The Head of Patient Safety & Improvement produces regular site and organisational level reports to highlight progress and outstanding actions at both site and organisational level. A review of outstanding actions will be undertaken with results reported back through the Patient Safety Strategic Group in September. Process Measure Process Measure WALK_NHSL Number of Patient Safety Leadership Walkrounds 70 60 50 Number 40 30 20 10 0 Jun-15 Number Goal: The goal is for a minimum of 1 Patient Safety Leadership Walkround to take place per week Description: This is a cumulated count of the number of Patient Safety Leadership Walkrounds that occur every month. Senior leaders are encouraged to use Patient Safety Leadership Walkrounds to demonstrate their organisation's commitment to building a culture of safety. Page 14 of 23
Safe: Closure of Category 1 rated incidents in Datix Safe: Closure of Category 2 rated incidents in Datix Closure of Category 1 rated incidents in Datix within agreed timescales Closure of Category 2 rated incidents in Datix within agreed timescales Total Category 1 rated incidents closed in Datix within 90 days Total Category 2 rated incidents closed in Datix within 30 days Total number of incidents in DATIX graded as Category 1 rated incidents 8 Category 1 incidents were recorded on Datix for April 2015. From the 8, 2 had SAER s commissioned. If incidents exceed the agreed timescale there is an escalation process involving operational Site Risk Management Facilitators and Senior Management Teams. A monthly review of Significant Adverse Events is undertaken by the Medical Director, Nurse Director, Corporate Risk Manager and Head of CG&RM with monitoring of the process for effective management of SAER S Total number of incidents in DATIX graded as Category 2 rated incidents The following chart sets out NHSL performance against category 2 incidents and timescales for closure of incidents. For incidents exceeding the agreed timescale there is an escalation process involving operational Site Risk Management Facilitators and Senior Management teams. Description: The risk management steering group set performance indicators (KPIs) The KPIs inform the board on the effectiveness of incident management. The board should look for performance to be improving. Page 15 of 23
Safe: Compliance with Hospital Length of Stay Average Hospital Length of Stay: Surgical Total bed days used for patients discharged from a Surgical specialty Number of hospital discharges, for patients discharged from a Surgical specialty The average Hospital Length of Stay (surgical) for June 2015 has increased in comparison to the previous month and is now showing normal variation. May 2015 data presented a special cause data point and was the lowest surgical ALOS in the two year reporting period. Dec 2014 data presented a special cause data point; poor flow and an increase in delayed discharges was noted in Dec 14 for NHSL. Further review of the data at hospital level indicates that the Surgical ALOS was particularly high for Hairmyres Hospital in Dec 2014. Average Hospital Length of Stay: Medical Total bed days used for patients discharged from a Medical specialty Number of hospital discharges, for patients discharged from a Medical specialty The average Hospital Length of Stay (Medical) for June 2015 has increased in comparison to the previous month. Dec 2014 data presented a special cause data point; poor flow and an increase in delayed discharges was noted in Dec 14 for NHSL. With the exception of December 2014, this data is showing normal variation Description: Reducing Length of Stay releases capacity in the system, including beds and staff time. This increase in capacity will help to minimise waiting times, maximise productivity and improve the patient experience. However, if patients are discharged too early this could lead to readmissions. Work is underway to identify a suitable benchmark as the overall Scottish average rate is not comparable. This will be incorporated at a future date. Description: Reducing Length of Stay releases capacity in the system, including beds and staff time. This increase in capacity will help to minimise waiting times, maximise productivity and improve the patient experience. However, if patients are discharged too early this could lead to readmissions. Work is underway to identify a suitable benchmark as the overall Scottish average rate is not comparable. This will be incorporated at a future date. Page 16 of 23
Safe: Compliance with Emergency Medical Readmissions Rate of Emergency Medical Readmissions within 7 days (per 1000 discharges) Number of emergency readmissions to any medical specialty within 7 days of discharge for patients initially admitted to a medical specialty Number of hospital discharges, for patients admitted to a medical specialty The 7 day Medical Readmissions Rate reporting has been moved into an SPC chart. The SPC chart displays non-random variation for the Dec 14 to Apr 15 period, and Dec 14 / Jan 15 are outwith the upper control limits - the increased rate appears to be driven by readmissions through Monklands Hospital and may reflect an issue with classification of admission type in ambulatory care. Work is underway to review this. The rate for May 15 has decreased in comparison to the previous month, and the data is now displaying normal variation. Rate of Emergency Medical Readmissions within 28 days (per 1000 discharges) Number of emergency readmissions to any medical specialty within 28 days of discharge for patients initially admitted to a medical specialty Number of hospital discharges, for patients admitted to a medical specialty The 28 day Medical Readmissions Rate reporting has been moved into an SPC chart. The SPC chart displays non-random variation for the Oct 14 to May 15 period, and Dec 14 / Jan 15 are outwith the upper control limits. - the increased rate appears to be driven by readmissions through Monklands Hospital and may reflect an issue with classification of admission type in ambulatory care. Work is underway to review this. READMM7_NHSL Medical Readmissions within 7 Days READMM28_NHSL Medical Readmissions within 28 Days Rate 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 Jun-13 Rate Median National Average Rate 160 155 150 145 140 135 130 125 120 115 110 105 100 95 90 85 80 Jun-13 Rate Median National Average Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median has been calculated using Sep 12-Aug 13 data points. National Average: Rate of readmissions should be benchmarked, with the goal of having a rate less than the national average rate of readmissions. The national average goal line is based on the last reported data on the national Hospital Scorecard (Sept 2014). See notes in Appendix A. Description: Readmissions within 7 days give a more accurate picture of readmissions which are clinically related to the index episode and it has been shown that readmissions within 7 days are likely to contain a higher proportion of 'avoidable' readmissions than the broader category of readmissions within 28 days Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median has been calculated using Sep 12-Aug 13 data points. National Average: Rate of readmissions should be benchmarked, with the goal of having a rate less than the national average rate of readmissions. The national average goal line is based on the last reported data on the national Hospital Scorecard (Sept 2014). See notes in Appendix A. Description: Around 15% to up to 20% of 28-day readmissions could be regarded as avoidable*. The most promising interventions to prevent readmission appear to be those that concentrate on coordination and communication around the time of discharge. *Preventing emergency readmissions to hospital, a scoping review, E. Nolte, M. Roland, S. Guthrie, Laura Brereton, 2012 (prepared for the UK Department of Health) Page 17 of 23
Safe: Compliance with Emergency Surgical Readmissions Rate of Emergency Surgical Readmissions within 7 days (per 1000 discharges) Number of emergency readmissions to any surgical specialty within 7 days of discharge for patients initially admitted to a surgical specialty Number of hospital discharges, for patients admitted to a surgical specialty The 7 day Surgical Readmissions Rate reporting has been moved into an SPC chart. The data is currently displaying normal variation. Rate of Emergency Surgical Readmissions within 28 days (per 1000 discharges) Number of emergency readmissions to any surgical specialty within 28 days of discharge for patients initially admitted to a surgical specialty Number of hospital discharges, for patients admitted to a surgical specialty The 28 day Surgical Readmissions Rate reporting has been moved into an SPC chart. The data is currently displaying normal variation. READMS7_NHSL Surgical Readmissions within 7 Days READMS28_NHSL Surgical Readmissions within 28 Days Rate 50 45 40 35 30 25 20 15 10 5 0 Jun-13 Rate 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 Jun-13 Rate Median National Average Rate Median National Average Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median has been calculated using Sep 12-Aug 13 data points. This has been recalculated due to a sustained shift, based on the 12 month period Nov 13-Oct 14 inclusive. National Average: Rate of readmissions should be benchmarked, with the goal of having a rate less than the national average rate of readmissions. The national average goal line is based on the last reported data on the national Hospital Scorecard (Sept 2014). See notes in Appendix A. Description: Readmissions within 7 days give a more accurate picture of readmissions which are clinically related to the index episode and it has been shown that readmissions within 7 days are likely to contain a higher proportion of 'avoidable' readmissions than the broader category of readmissions within 28 days (Levy et al, 2000).**Clinical Outcome Indicators, Clinical Resource and Audit Group (CRAG), 2002 Median: The median enables run chart rules to be applied to identify if there has been a non-random change. (Please see page 17) The baseline median has been calculated using Sep 12-Aug 13 data points. This has been recalculated due to a sustained shift, based on the 12 month period Nov 13-Oct 14 inclusive. National Average: Rate of readmissions should be benchmarked, with the goal of having a rate less than the national average rate of readmissions. The national average goal line is based on the last reported data on the national Hospital Scorecard (Sept 2014). See notes in Appendix A. Description: Around 15% to up to 20% of 28-day readmissions could be regarded as avoidable*. The most promising interventions to prevent readmission appear to be those that concentrate on coordination and communication around the time of discharge. *Preventing emergency readmissions to hospital, a scoping review, E. Nolte, M. Roland, S. Guthrie, Laura Brereton, 2012 (prepared for the UK Department of Health) Page 18 of 23
Quality Ambition: Effective "The most appropriate treatments, interventions, support and services will be provided at the right time to everyone who will benefit, and wasteful or harmful variation will be eradicated." Progress on this ambition is measured through clinical quality indicators, for stroke care and A&E waiting times Effective: Admission to Stroke Unit & Stroke Treatment Indicators Admission to Stroke Unit No. of patients admitted to stroke unit on day of admission or day after. Total no. of inpatients with diagnosis of stroke. All targets met NHSL wide for Stroke Unit admission in June STRKUNIT_NHSL 100 Admission to Stroke Unit STR_TRE 100 Stroke Treatment Indicators CT: No. of patients who had CT/MRI within 24 hours of admission. Swallow Screen: No. of patients who had swallow screen on day of admission. Aspirin: No. of patients who had aspirin on day of admission or day after. Total no. of inpatients with diagnosis of stroke. All targets met NHSL wide for Swallow screen and CT imaging. 2 Aspirin failures, due to 1 patient with initial query re intracranial bleed, & 1 patient on end of life care, with aspirin not given earlier for good medical reason, both at Wishaw. On checking data, aspirin omitted on essca for 2 patients by the time snapshot was taken, now rectified, closer monitoring being maintained. Monklands did not achieve target for swallow screening, 2 patients too unwell on admission, 1 of these being treated immediately by Cardiology, and 2 patients arriving at A&E late in the evening, however both received swallow screening after midnight. Process Measure Stroke Treatment Indicators 80 90 % Compliance 60 40 % Compliance 80 70 60 20 0 % Compliance Goal Jun-15 50 40 Swallow CT Aspirin CT/Swallow Target Aspirin Target Jun-15 Goal: Scottish Stroke Care Standards - 90% of all patients admitted to hospital with a diagnosis of stroke are admitted to the stroke unit on the day of admission, or the day following presentation at hospital. Description: The Scottish Stroke Care Standards focus on those parameters which have the best evidence for having an effect on patient outcomes. Data reliability = High (see appendix c) Goal: Scottish Stroke Care Standards - 90% of patients have CT/MRI imaging within 24 hours of admission / 90% of patients have swallow screen on day of admission / Aspirin given on day of admission or following day for 95% patients (changed from 100% in January 2015). Description: The Scottish Stroke Care Standards focus on those parameters which have the best evidence for having an effect on patient outcomes. Data reliability = High (see appendix c) Page 19 of 23
Effective: Compliance with A&E Treatment Target % of A&E attendances waiting more than 4 hours The number of patients waiting for more than 4 hours at an A&E clinic. The number of patients attending an A&E department. Four hour waiting time breaches in June have decreased in comparison to the previous month. Compliance with this metric has not been within target since July 2012. The Acute Division are reviewing plans to significantly reduce these waiting times and more detailed narrative will be incorporated in future reports. % of A&E attendances waiting more than 12 hours The number of patients waiting for more than 12 hours at an A&E clinic. The number of patients attending an A&E department. 12 hour waiting time breaches for June have decreased in comparison to the previous month. February's rate was the highest in the two year reporting period at 1.33%. Compliance with this metric has not been within target since August 2013. The Acute Division are reviewing plans to significantly reduce these waiting times and more detailed narrative will be incorporated in future reports. AE4HRS_NHSL A&E Waiting Times Breaches - 4Hrs AE12HRS_NHSL A&E Waiting Times Breaches - 12Hrs 20.0 1.4 % of A&E Waiting times breaches 17.5 15.0 12.5 10.0 7.5 5.0 2.5 0.0 % breaches Goal Jun-15 % of A&E Waiting times breaches 1.2 1.0 0.8 0.6 0.4 0.2 0.0 % breaches Goal Jun-15 Goal: 5% or less of A&E attendances waiting more than 4 hours Description: The Scottish Government standard for Accident and Emergency departments is that 95 % of all attendances are seen within 4 hours. The board should look for % of A&E waiting times breaches to be at or below the compliance line. Goal: Eliminate 12 hour waits Description: Actions against AIM 9 in the 2014-17 NHSL Quality Assurance & Improvement Strategy includes measuring and reducing patient harms relating to patient flow. The Board should look for % of A&E waiting times breaching 12 hours to be reducing to zero. Page 20 of 23
Appendix B - Notes on the data Average Hospital Length of Stay. The data reported only shows Inpatient Hospital Discharges. Total Hospital Length of Stay (HLOS) is linked to the discharge specialty and discharge date even if part of the hospital stay took place under different specialties and/or across different specified dates. This ultimately means some outlier long stays could have a big impact on Average HLOS for a specific specialty. The data reported only shows admissions where hospital Spell start date = admission date, therefore some hospital transfers will be excluded. Measures have not been standardised by age, sex and deprivation Average HLOS has not been adjusted for case mix so may reflect variations in complexity of patients treated Denominators are based on discharge dates to allow shorter reporting lag times. This means that the data on the local report will not be directly comparable with the national Early Warning Scorecard which uses admission dates for reporting. Data are subject to change following SMR validation processes. Readmissions d ata Me asures have not b ee n standardised by age, sex and deprivation R ates are only calculated using linked data held in the La narkshire TrakCare PMS and therefo re doe s not include re admissions to other Health Boards D enominators are based on discharge dates to allow shorter reportin g lag tim es. This m eans that the data on the local re port w ill not be directly compa rab le w ith the national Early W arning Scorecard which uses admission da tes for repo rting. D ata are subject to change following S MR validation processes. Deaths and Live Discharges The denominator for Unadjusted Patient Mortality includes Deaths and Live Discharges in all areas except for Obstetrics. Both the numerator and the denominator for Cardiac Arrest Rate exclude Cardiac Arrests and Deaths and Live Discharges in these areas respectively: - -Hairmyres CCU -Hairmyres Day Surgery Unit -Hairmyres Dental Day Unit -Hairmyres ITU -Hairmyres Maternity Day Assessment -Monklands Endoscopy Unit -Monklands Hospital Day Surgery Unit -Monklands Ward 16 Haem atology Day Unit -Monklands Ward 26 ITU -Wishaw General CCU -Wishaw General ICU -Wishaw General Medical Day Unit -Wishaw General Neonatal ICU -Wishaw General Surgical Day Unit -Wishaw General Ward 19 -Wishaw General Ward 20 -Wishaw General Ward 21 -Wishaw General Ward 22 -Wishaw General Ward 23 -Wishaw General Ward 24 Page 21 of 23
Appendix B - Notes on the data (continued) Interpreting run charts using run chart rules RULE 1: A SHIFT 6 or more consecutive points either all above or all below the median line RULE 2: A TREND 5 or more consecutive points all going in the one direction (up or down) RULE 3: TOO MANY OR TOO FEW RUNS A non random pattern would be indicated by the data crossing the median too many or too few times (Reference table for this one) RULE 4: AN ASTRONOMICAL VALUE A value that is obviously, blatantly different and really stands out as being highly unusual Data Reliability Each metric has been graded for data reliability against the Clinical Quality Data Reliability Matrix (see Appendix C). The Data Reliability score is calculated based on the data source and sample size vs. the highest level of validation the data has been subject to. Notes: Whilst there is a drive to provide Clinical Quality data that is as complete, accurate and valid as possible, it is also important that data for local improvement is available as real time as possible and full validation is therefore not always possible or necessary. Clinical Quality data is reported from a variety of sources. The method of data collection and source of data will have an impact on the level of data validation, and subsequently the reliability of data reported: For example, data collected via direct observations of clinical practice (e.g. patient safety process measures) are based on an individual s judgement that something has happened at a particular point in time. This data can never be checked against source data (e.g. clinical records) and there is a reliance on the individual to provide an accurate account of events. Observational data would therefore be considered of low reliability in terms of accuracy. However, since observational measurement is in the main data for improvement, this level of reliability would be acceptable. Conversely, data extracted from clinical systems (e.g. laboratory results) should be of high reliability in terms of accuracy, as this is clinical data used for managing patient care. For reporting purposes however, this type of data would still need to be checked for format, completeness, duplicates and outliers and the level of reliability would increase depending on the level of validation applied. Page 22 of 23
Appendix C - Clinical Quality Data Reliability Matrix HIGH Reliability of data LOW Reliability of data HIGH DATA SOURCE 5 4 3 2 Higher score = greater level of confidence in reliability of data Data extract from clinical systems (PMS / Labs etc.) Data transcribed from clinical records onto paper forms / databases (100% sample) Data which relies on individuals reporting incidences (e.g. extract from DATIX / cardiac arrest) Data from surveys / questionnaires / small sample casenote reviews HIGHEST LEVEL OF VALIDATION 5 4 3 2 1 Validation checks on format, External QA Locally validated completeness, Sense checks against source against source duplicates, prior to reporting data (e.g. ISD) data outliers & No validation standards 25 20 15 10 5 20 16 12 8 4 15 12 9 6 3 10 8 6 4 2 LOW 1 Data from direct observations of clinical practice 5 4 3 2 1 Page 23 of 23