The results of a rheumatology case-mix survey has highlighted a need to better evaluate care pathways for patients with chronic diseases. Sangeetha Baskar and colleagues explain the evidence.
Increasing emphasis is being placed on the management of patients with chronic diseases such as rheumatoid arthritis in the community. There is also a perception that many such patients are followed up in specialist clinics unnecessarily and for protracted periods.
Attempts to tackle this problem have been crudely based on individual units’ new to follow up ratios. Such an approach has been criticized for not taking into account the case-mix of patients in each unit or the relevant disease management guidelines.
Case-mix describes and measures clinical activity by assigning patients to discrete diagnostic categories. It allows constructive inferences at local level and the study of differences in structure and provision of care at a regional and national level.
The aim of this study was to evaluate out-patient new to follow-up ratios (NTFUR) of individual rheumatology units across the Midlands. We also aimed to study the influence of various factors including case-mix in the inter-unit variation in NTFUR.
A West and East Midlands out-patient survey was conducted over a four-consecutive week period in March-April 2009. Patients were grouped into five diagnostic categories based on the final diagnosis:
- Rheumatoid arthritis (RA)
- Other Inflammatory arthritis (OIA)
- Connective tissue diseases including vasculitis (CTD)
- Non inflammatory arthritis (NIA) including osteoarthritis, osteoporosis, crystal arthropathy, soft tissue diseases, non inflammatory back pain and fibromyalgia
- Diagnosis not yet known (NK).
The first three conditions were classed as inflammatory (and hence appropriate for specialist follow up as per national guidelines) and compared against the non-inflammatory condition prevalence (potentially dischargeable) in the new and follow up settings.
The total number of patient contacts was 10,343; 7825 (76 per cent) from 13 West Midlands units and 2,518 (24 per cent) from four East Midlands units (range 343 -1089 /unit; average 608). There was a wide variation in the number of consultants (range 2-9), other consulting doctors (4-14) and clinical nurse specialists (2-9) between units.
Of all patient contacts, 17 per cent were new and 83 per cent were follow-ups giving an overall NTFUR of 1:4.9. The breakdown of various diagnoses of RA, OIA, CTD, NIA and NK arrived at the end of the consultation respectively were 17 per cent, 11 per cent, 10 per cent, 50 per cent and 12 per cent among new patients and 54 per cent, 16 per cent, 15 per cent, 13 per cent and 2 per cent among follow-ups.
There was a wide variation in the NTFUR between the units (range 1:3 to 1:7.3) and figure 1 summarises the diagnostic case mix and the NTFUR of the individual (anonymous) units.
Case-mix was analysed in relation to the difference in NTFUR between units (figure 2). Units were categorised based on their NTFUR’s to equal tertiles with the lower and higher tertile groups respectively indicating units with lower and higher number of follow-ups per new patient.
The proportion of RA and OIA was similar across the three groups and the significant difference (p<0.001) in case mix was driven by the proportion of non-inflammatory conditions and connective tissue diseases.
Stepwise linear regression analysis confirmed the diagnostic case mix to be the leading significant independent factor predicting the variance in NTFUR (adjusted R2 0.40, Beta 0.63, F 6785, p<0.001). Additional significant variables included number of consulting doctors (adjusted R2 0.57, Beta -0.44, F 7117, p<0.001) and number of Nurse Specialists (adjusted R2 0.58, Beta -0.04, F 4763, p<0.001) in each unit, more consulting staff correlating with greater new patient capacity.
Among patients with inflammatory conditions, 92 per cent and 98 per cent were given further appointments in the new and follow up settings respectively while 50 per cent of all patients with non-inflammatory conditions were routinely discharged at their first visit and another 36 per cent discharged at their follow up visit.
Significant variation in the outpatient NTFUR between units was accounted for in large part by underlying case mix. Units with lower non-inflammatory and higher inflammatory condition prevalence had higher number of follow-ups per new patient, indicating appropriate specialist follow up of inflammatory rheumatic conditions.
While the non-inflammatory condition prevalence in the new patient setting seemed high, it is important to recognise that the diagnosis of NIA was arrived at the end of the consultation often for patients referred as suspected inflammatory arthritis. The high discharge rate of patients with NIA both in the new and follow-up settings confirmed appropriate use of specialist resources.
Smaller studies have found similar influence of case mix on NTFUR. Bamji et al observed discordant NTFUR (1:2.1 and 1:4.2) between two neighbouring rheumatology units and found the presence of a physiotherapy-led back pain triage service in one unit contributed to a reduction in the number of non-inflammatory new patients.
Surveys in the South West UK also showed a significant variation in the NTFUR between rheumatology units and additionally found a trend in increasing follow-up workload with time. Analysis of dermatology outpatient services throughout Wales highlighted the need to avoid an inappropriate ‘one fits for all’ NTFUR and the need to benchmark NTFUR against diagnosis.
Comparison of case-mix with equivalent services in other trusts provides several benefits. It allows appropriate allocation and justification of use of existing resources, with greater sophistication than gross NTFUR. Considerable difference between units may be additionally explicable, for instance by the presence of nearby specialist units, or a difference in the relationship with other specialities in the local area.
Alternatively, it may highlight a shortfall of activity in a particular area or a surplus of non-inflammatory conditions that is possibly more appropriately dealt with either in primary care or by a multidisciplinary triage service.
Our findings highlight the dangers of uncritical comparison of crude outpatient new to follow up ratios as a basis for planning care and the need to include case mix and other local factors in any comparative exercise. These findings will be of value in commissioning future development of rheumatology services.
The authors are writing on behalf of the West Midlands Rheumatology Services and Training Committee and the East Midlands Regional Audit Group.