Referral-to-Treatment for Knee Arthroscopies A Report from the Musculoskeletal Audit Interpretive text from Colin Howie (Consultant Orthopaedic Surgeon, Royal Infirmary Edinburgh; Chairman, Scottish Committee for Orthopaedics and Trauma) To achieve 18 weeks from referral to surgery for patients undergoing knee arthroscopy (the commonest procedure carried out in orthopaedics) we need to understand the pressures and pinches in the system. No one solution will fit all, and it is striking that there is considerable regional variation in the patient journey across the country. In some cases simple changes in process could result in a significant shortening in the pathway. Here we report on a 13-week arthroscopy audit period from 14th September to 11th December 29. We included all patients listed for an elective knee arthroscopy. MSk Local Audit Co-ordinators collected data from patient case notes, patient information systems, results reporting and referral management systems. This report is not intended to assess hospitals and waiting times, and does not take into account stops that are beyond a hospital s control (e.g. delays caused by patient preference). It is intended as a tool for management, to help managers and clinicians identify areas where Referral to Treatment delays are occurring and/or may be minimised. Please send comments and queries to either Jane Campbell 1463 7585 MSk Clinical Co-ordinator jane.campbell7@nhs.net or Rik Smith 131 275 74 MSk Senior Analyst rsmith11@nhs.net
How to use this report The Referral to Treatment period of knee arthroscopy patients is summarised on an individual hospital basis, firstly (A) for the whole period from Referral to Treatment, then in more detail for the constituent parts of this process: B Referral to Outpatient Appointment (OPA) C OPA to Diagnostics/Review D Diagnostics E Surgery Note that not all patients followed a conventional path from referral to outpatient appointment, subsequent diagnostics and review to surgery (patients were referred directly for MRI from review following a previous knee or other treatment, were referred into the system to an NHS hospital from a private clinic, or were scheduled for a knee arthroscopy within a short time of attending the Emergency Department). Also, some dates for various parts of the patient journey (particularly vetting and allocation dates) were not always available. This explains variation in sample sizes across the report. The 18-week Referral to Treatment target does, of course, allow for clock-stopping for circumstances beyond a hospital s control. Allowing for all such patient-centred delays was beyond the remit of this audit, although some summary data are included. Inspection of the constituent parts of the patient referral-tosurgery period should reveal the main limiting steps in each unit, but, if it would be useful, please ask us to select or break down your information further. During the audit period the Local Audit Co-ordinators achieved a near 1% inclusion rate of knee arthroscopies for all participating hospitals (Table 1). In some hospitals, however, a small number of patients who were eligible for inclusion in the audit had to be excluded due to a lack of comprehensive or complete data, or case notes being unavailable to the MSk Local Audit Co-ordinators. Table 1: Number of patients reported on Hospital Number of arthroscopy patients identified in report period in each hospital Number of arthroscopy patients included in this report Inverclyde/Dunoon 82 79 37 37 46 45 143 14 * 55 55 46 44 174 174 76 76 cathro ** 98 98 33 33 29 29 17 17 23 23 236 23 133 128 82 82 Total 14 138 Patients are mostly grouped by surgical hospital, except for: * Includes 3 patients originating from that were operated on at. ** Also includes 38 patients originating at Ninewells who were operated on at PRI - - 2
A) Summary of overall period from Referral to Treatment Musculoskeletal Audit (MSk) - Arthroscopies Fig. 1: Length of time from referral to surgery - all patients 1 8 6 4 2 <=18 weeks 18-24 weeks >24 weeks Missing data Missing data if referral date was not known, often from an unconventional source (e.g. referral from a review of treatment on the other knee, back pain), or straight to MRI or listing for surgery Fig. 1 includes 417 (3%) patients who had indicated that they were unavailable at some stage between referral and surgery. If unavailable at some stage, documented unavailability averaged 39 days (median=28 days), and 128 (9%) patients were known to be unavailable for more than six weeks. Only 8% of patients who had indicated that they were unavailable at some stage between referral and surgery were treated within 18 weeks, compared to 31% of patients who were available throughout. Fig. 2 reports the percentage of patients who met the 18-week Referral to Treatment target after excluding these unavailable patients. Fig. 2: Length of time from referral to surgery - excluding patients who were unavailable at some stage during this period 1 8 6 4 2 <=18 weeks 18-24 weeks >24 weeks Missing data - - 3
Altogether 39% of all patients had some patient-centred delay (unavailability; cancelled OPA, review or surgery; did not attend OPA, review or surgery; declined appointment; medically unfit for theatre; Fig. 3). These patients took a median of 3 weeks from referral to surgery, compared to 22 weeks in patients with no documented patient-centred delay. Fig. 3: with some documented patient-centred delay (see text for definition) 1 8 6 4 2 Surgeons regularly complain that there are gaps in clinics. Although we aspire to treat all patients within 18 weeks, a substantial number of patients fail to attend for whatever reason. It is interesting that these patients not only incur longer waits but also that unavailability rates vary widely around the country. Note that the above figure will not show all patient preference delays. For example, some clinics/waiting lists may phone patients to offer a choice of dates for OPA or surgery, and the only data available to the MSk Audit was the date accepted. - - 4
Although it would appear that most patients are initially seen within 9 weeks (Fig. 4) this excludes the diagnostic pathway (MRI) and subsequent clinic visits. Should the target for Outpatient Appointment be shorter to help accommodate these delays or can we reduce delays by wider use of Instant MRI? Fig. 4: Average time spent in each stage of the Referral to Treatment process - all patients Mean number of weeks 4 3 2 1 From referral to OPA Diag/review to theatre list OPA to diagnostics/review Theatre list to surgery NOTES Times are means for each stage for patients whose data was available Time from OPA to diagnostics/review is to the first review clinic, or the date the diagnostic (MRI) was performed if the patient was reviewed by phone or letter Time from diagnostics/review to the date the patient was added to the theatre list is time from first review, and may have included subsequent reviews/diagnostics If patients were added to the theatre list at the initial OPA, and were not referred for diagnostics or a subsequent review (approximately half of all patients), OPA to diagnostics/review was included as weeks, and diagnostics/review to theatre list as the time between the initial OPA and the date the patient was added to the theatre list. A small number of patients had diagnostics/reviews after they were apparently added to the theatre list. These were excluded from calculation of time from diagnostics/review to theatre list The mean time for OPA to diagnostics/review includes many patients who were added to the waiting list for surgery at the initial OPA (i.e. time from OPA to diagnostics/review= weeks). Therefore the actual distribution of times from OPA to diagnostics/review is not reflected by the mean, and may be skewed by a small number of long waiting times. - - 5
B) Referral to Outpatient Appointment 96% of 132 patients with documented length of time from Referral to Outpatient Appointment were seen within 15 weeks (Fig. 5). Fig. 5: Length of time from Referral to Outpatient Appointment 1 8 6 4 2 <=5 weeks 5-1 weeks 1-15 weeks >15 weeks Missing data Fig. 5 includes all patients irrespective of any reason for delay, and reflects the 15-week target for Referral to Outpatient Appointment in operation during the audit period. Comparison against the 12-week target times which come into place in March 21 are available on request Although it was usually possible to find out who vetted patients and the vetting outcome, in many hospitals first vetting date was often unavailable (Table 2). This may improve with the introduction of e-vetting in some sites. Amongst patients whose vetting times were known, 93% of patients were vetted within a week (Table 2). Table 2: Average time from Referral to First Vetting % of patients whose vetting dates was documented Number of patients documente d Referral to vetting Mean (weeks) % > 1 week 65% 5.6 4% 11% 4. % 89% 39. % 37% 48.9 19% 65% 36.7 8% 5% 22.5 9% 1% 1. % 55% 42.7 17% cathr o 16% 16.2 6% 73% 24. % 3% 1. % 5% 53.1 % 87% 2.2 % 86% 196.3 4% % - - 24% 2 1.7 3% 42% 572.4 7% - - 6
Hospital vetting policies vary considerably (Fig. 6). If known, almost all patients originally vetted by clerical staff were subsequently vetted by medical staff. The overall time from referral to surgery was shorter if patients were first vetted by medical staff compared to ESPs or clerical staff (median 23 weeks compared to 27 weeks). When known, 83% of patients who were vetted twice were second-vetted within a week and all patients were second-vetted within three weeks. Two hospitals ( and ) averaged more than a week to second vetting. Fig. 6: Who first vetted patients post Referral? 1 8 6 4 2 Medical staff ESP Clerical staff Not vetted Missing data 64% of patients who were not vetted were unscheduled referrals (e.g. from ED). Others included review patients going on to have a second procedure. As expected, much of the time elapsed between Referral and Out Patient Appointment occurred whilst the patient awaited their offered OPA date. The average wait from the date the OPA was ready for allocation to the date offered for OPA was 5.5 weeks, but many hospitals had a considerable proportion of patients who waited more than eight weeks (Table 3). Table 3: Average times from Vetting to Outpatient Appointment Vetting to Ready for Allocation N Mean (weeks ) % > 8 week s N Allocation to Date Offered Mean (weeks) % > 8 weeks Date Offered to OPA attendance N Mean (weeks ) % > 8 week s 49 2.3 % 69 3.1 1% 73.6 3% 4. % 36 7.5 53% 36.9 6% 4.6 % 44 8.1 61% 45.1 % 5 5.6 16% 99 3.5 3% 13.3 2% 36 6.8 39% 55 2.8 2% 55.4 2% 22.4 5% 41 7.2 54% 41.1 % - - 131 4.4 11% 16.2 1% 42.1 % 74 6.5 49% 76.2 % cathr o 16.8 6% 82 4. 1% 94.5 2% 24. % 27 6.5 37% 32.1 % 1. % 28 7.1 46% 29. % 53.4 % 91 6.9 46% 1.3 % 19.1 % 2 7.5 5% 22.4 % 193 1.7 1% 27 6.5 37% 216.7 3% - - 14 6. 34% 128.5 2% 17 3.8 12% 59 3.6 2% 73.3 % 566 1.9 5% 1167 5.5 27% 131.4 2% - - 7
Seven per cent of OPAs were delayed because patients either Did Not Attend, cancelled or were otherwise unavailable (Fig. 7). A further 1% were delayed because the hospital cancelled the appointments. The high percentage of patients waiting more than 8 weeks suggests that departments may need to increase clinic capacity. Although the actual number of patients involved is small, this is only a subset of patients referred for an opinion. Fig. 7: First reason for delay to OPA 3 2 1 No delay Unknown reason Patient unavailable (includes 'declined appt') Clinic cancelled original appointment Patient cancelled original appointment Did Not Attend Excludes 7 patients whose offer date for OPA was unknown, or who referred directly for MRI or surgery without a conventional OPA (e.g. after review clinic for previous knee treatment, from private clinics for NHS treatment, etc) There are surprising differences around the country which suggest differences in coding and practice. Some providers have no patients unavailable, while others have large volumes. - - 8
C) Time from Outpatient Appointment to Diagnostics/Review Musculoskeletal Audit (MSk) - Arthroscopies Knee arthroscopy patients were managed in a number of different ways following their initial Outpatient Appointment (Fig. 8). 45% were added to the waiting list for surgery without further diagnostic tests or further review. 47% were sent for MRI scans (only rarely for other diagnostic tests), and were then either reviewed by further OPA or by phone/letter. The remaining 8% were referred for continuing management, treatment or to other clinicians, and were reviewed by further OPA prior to surgery (Fig. 8). The variation in practise around the country is extensive, and merits a detailed national evaluation. Fig. 8: Path from Outpatient Appointment to Surgery 1 8 6 4 2 Added to surgical waiting list at OPA MRI - reviewed by phone/letter Other MRI - reviewed by further OPA No MRI, but further review OPA Of those that were sent for further diagnostics or review 67% were reviewed within 12 weeks (Fig. 9). Fig. 9: Length of time from Outpatient Appointment to Diagnostics/Review 1 8 6 4 2 <=6 weeks 6-12 weeks 12-18 weeks >18 weeks Missing data Time is from OPA to first review OPA, or to date of MRI if patient was reviewed by phone/letter. Excludes patients who had neither an MRI nor a further review - - 19
Fig. 1 clearly shows that reviewing patients in outpatients prior to listing does increase delay. Indeed, taken together, Figs. 8, 9 and 1 suggest some areas where hospitals could revise practise to reduce their waiting times. Some hospitals send many patients for MRI yet the waiting time is short, a classic capacity versus demand issue? Perhaps agreed national guidelines could be helpful. Fig. 1: Average time from OPA to Review in relation to diagnostics and review method Average time (weeks) to review 2 18 16 14 12 1 8 6 4 2 No MRI, but further review OPA MRI, reviewed by phone/letter MRI, reviewed by further OPA Time is to first review or to date of MRI if reviewed by phone/letter. If there were fewer than three patients in a particular category in Fig. 1, the data is not shown If a further review appointment was made, 86% of patients with known planned review dates were reviewed on that date. 67 (14%) patients who attended later attended a mean of 4.5 weeks later. Most delays were due to patients who cancelled, were unavailable or Did Not Attend, but 19 patients were cancelled by the clinics. Eleven per cent of all patients had more than one review clinic (2% of all patients who had diagnostics and/or subsequent review). Of the 114 patients known to have two reviews, mean time between reviews was 11 weeks (median 9 weeks). 4 patients had more than two reviews and a mean time of 32 weeks between first and last reviews (median 22 weeks). - - 1
D) Diagnostics The average time from the date an MRI was ordered to the date it was carried out varied from two to eight weeks across hospitals. The distribution of waiting times for MRI is shown in Fig. 11. Fig. 11: Time from date of MRI request to MRI performed: 1 8 6 4 2-2 weeks 2-4 weeks 4-8 weeks >8 weeks Missing data If patients had an MRI and were then reviewed by a further Outpatient Appointment, the average time from MRI to the review OPA was 6 weeks (median 5 weeks), and this varied between 3 and 8 weeks between hospitals. 38% of patients had their review OPA within 4 weeks of the MRI, but this varied from 16% to 77% between hospitals. - - 11
E) Surgery 84% of patients were operated on within 15 weeks of being added to the theatre waiting list (Fig. 12), but many of these were in the period from 1-15 weeks, a substantial chunk of the future 18-week RTT target period. During the audit period the target time for the whole period from diagnostic to surgery was 15 weeks, and this will reduce to 12 weeks in March 21. Fig. 12: Length of time to surgery after being added to the theatre waiting list 1 8 6 4 2 <=5 weeks 5-1 weeks 1-15 weeks >15 weeks Missing data Fig. 12 includes all patients irrespective of any reason for delay. Table 4: Average component times from Decision for Surgery to Date of Surgery Time from last OPA/diagnostic to date added to theatre waiting list* Time from being added to theatre waiting list to first theatre date booked % not operated on at first booked theatre date Time from first date offered to actual theatre date N Mean % > 8 Mean % > 8 Mean % > 8 N N (weeks) weeks (weeks) weeks (weeks) weeks 79.7 3% 78 12.3 8% 15% 78 1. 4% 37.7 3% 37 9.9 59% 5% 37.1 % 44.7 2% 44 8.2 42% 4% 45.8 2% 137.4 1% 14 1.1 76% 9% 14.6 2% 55.4 % 55 8.9 64% 18% 55.8 2% 41.1 % 44 7.8 41% % 44. % 173.3 1% 173 1.8 74% 3% 174.5 2% 76.6 4% 76 9. 7% 18% 76 1.1 4% 97.4 1% 98 9.1 6% 17% 98 1.4 4% 31.3 % 33 12.9 64% 18% 33 2.2 9% 29.2 % 29 8.6 55% 7% 29 1.7 7% 11 1.8 7% 17 9.5 52% 4% 17.5 2% 23.7 4% 22 11.6 74% 14% 22 2.5 13% 223.9 4% 23 1.4 8% 21% 23 2.4 1% 127 1.3 5% 125 8.8 57% 6% 126.5 3% 78.8 5% 82 8.4 59% 18% 82 1.6 5% Total 1351.7 3% 1373 9.8 67% 12% 1376 1.1 4% * Includes patients who went straight to theatre waiting list following their initial Outpatient Appointment. Some patients were added to the theatre waiting list before their final diagnostics result was available and are included above as being added immediately. Table 4 includes all delays irrespective of any reason for delay. In some instances the delay may be due to the patient cancelling the theatre date, failing pre-assessment or being unfit for theatre on the day of admission (see Fig. 13) - - 12
On average 12% of all patients did not go to surgery on their original booked date (Fig. 13). Fig. 13: Reason for delay from first booked theatre date to actual date of operation 3 2 1 No delay Missing data Surgery cancelled following admission Failed pre-assessment Hospital cancelled Patient cancelled Did Not Attend Seven patients were cancelled on the day they were admitted because they were medically unfit, and five because of lack of theatre time, lack of equipment or equipment failure. Again there are considerable variations in causes of delay. Listing surgeons should ensure that patients are aware of the implications of listing and when they will be called for surgery. If the patient is not available they should not be listed. The majority of knee arthroscopy patients are now treated as daypatients (Fig. 14). Fig. 14: Surgery daypatient or inpatient and whether operated on day of admission 1 8 6 4 2 Missing data Inpatient - operated on day after admission Inpatient - operated on day of admission Daypatient - - 13
Fig. 15: Type of operation 1 8 6 4 2 Missing data or unable to code Other or combination Other soft tissue procedure Ligament repair Debridement cartilage trim/microfracture Menisectomy Diagnostic knee Other includes cartilage grafts (autologous or synthetic) - - 14