The primary aim of this study was to identify factors from GP referral letters that can predict which patients with knee and/or hip pain would receive an optimal care pathway at the time of consultation. Factors which were found to predict an optimal care pathway were: lower BMI,, having a named disease or syndrome and taking a pharmacologic substance. Having a single diagnostic procedure predicted a sub-optimal pathway. Over 30% of participants were found not to have had an optimal care pathway. The secondary aims were to identify predictors of patient rated pain and function at time of consultation and after 6 months and to describe the characteristics of the care pathway. Variables found to predict pain and function at initial consultation were higher age, higher BMI, current smoking, with knee pain, having sign or symptoms and having therapeutic/preventative procedure and opioid medication history. Of these variables only age, BMI, smoking status and medication history were individually found to predict pain and function at 6 months post consultation. Only predictors related to BMI were predictors of both optimal care and pain and function. A key characteristic of the care pathway for individuals with knee and/ or hip pain is that treatments received varied according to the type of specialist clinic seen in.
Predictors of receiving care on an optimal versus sub-optimal pathway at the time of initial consultation
Our findings suggest that not all patients received a treatment outcome that resulted in an optimal care pathway. This represents potential inefficiency and wasted healthcare resource use. In this particular cohort of patients this could be improved for over 30% of cases, which has not previously been quantified in the musculoskeletal literature. Variables associated with optimal care and predicted 10% of variance were lower BMI and three concepts from the free text of the referral: having a named disease or syndrome and taking a pharmacologic substance. Having a single diagnostic procedure predicted a sub-optimal care pathway. These factors should be routinely included in referrals as part of a minimum dataset. Despite BMI being a strong predictor it was frequently unreported, so addressing this in future referral guidelines is essential. Pain and function at time of consultation or at 6 months post consultation did not predict receiving optimal care. One explanation for this is that the definition of ‘an optimal pathway’ used in this study is about efficient resource allocation and does not consider patient rated pain and function. Therefore, efficient use of healthcare resources is not necessarily related to patient opinion of their condition and these are independent concepts.
Predictors of patient rated pain and function at time of consultation
Factors that predicted baseline patient rated KOOS and HOOS for combined pain and function across all participants regardless of clinic type or pathway were higher age, higher BMI, current smoking, with knee pain, having sign and symptoms, having therapeutic/preventative procedure and opioid medication history. Individually, many of these variables were also found to predict combined KOOS and HOOS scores at 6 months: age, BMI, smoking status, co-morbidity index and medication history. This should be interpreted with caution as this is not part of the multivariate analysis but all of these factors should be routinely documented in referrals for specialist opinion.
Based on the study findings it is apparent that factors used to predict optimal care are different to those that predict pain and function. Furthermore, BMI was a predictor for both receiving optimal care and pain and function outcome, it is therefore essential that this is included in any future minimal dataset.
The methods using in this study are novel for triaging referrals for specialist opinion. No previous studies have evaluated care factors that predict who received optimal care, but these factors do need to be considered in any future referral or triage system. When developing prioritization tools for patient triage it is essential to include demographic data and variables that we have identified from the free text component of the referral.
Previous studies have evaluated triage prioritisation tools for hip and knee pain, but these were not based on predictor variables [10] and there had been a lack of transparency about what variable prioritisation is based on [11, 25, 26]. Further research is required to validate the findings of this study and to develop the prioritisation tools and training required for an optimal pathway that could be tested in a randomized control trial in the future. This staged approach is in line with the IDEAL-Physio framework for guiding innovation and evidencing interventions [27].
Care pathway characteristics
A secondary aim of this study was to describe the characteristics of the care pathway for hip and / or knee pain at the point of referral for specialist assessment according to specialist clinic type and receiving care on an optimal or sub-optimal pathway. The care pathway that patients in this study followed is displayed in Fig. 3. The organisation of care around three different professional specialities (specialist GP, advanced physiotherapy practitioner and orthopaedics) represents further variation compared to that already described in the literature. For example, combined physiotherapist and orthopaedic clinics [28,29,30,31,32], separate orthopaedic and physiotherapist clinics [25, 33] or musculoskeletal clinical assessment triage and treatment service (MCATS) combing advanced physiotherapy practitioner and orthopaedic consultant, physiotherapy led clinics [34] separate Orthopaedic clinics [7].
In the current study the treatment outcomes are reported according to the clinic type. There was higher referral rate to conservative treatments and imaging by advanced physiotherapy practitioner, higher injection rate by specialist GP and higher rates of surgical intervention for participants seen in orthopaedic clinic. There is a scarcity of evidence in the literature around treatment outcomes for the different care pathways for hip and knee pain. Data that are available also suggest that patients seen in a physiotherapist or musculoskeletal care, assessment and treatment clinic are more likely to receive an injection, non-steroidal anti-inflammatory drug prescription, a course of physiotherapy or conservative treatment [6, 7, 32]. Therefore, there is a risk that there will be variation in treatment offered based on the professional background of the healthcare professional [7]. Furthermore, it could be argued that these patients should receive conservative treatments in primary care, before being referred for specialist opinion and represents inappropriate referral [1].
In addition, individuals seen in the advanced physiotherapy practitioner clinics tended to be younger, have lower BMI, have fewer co-morbidities, take fewer medication, have a higher level of function and less pain. This would seem to corroborate the finding that this group of individuals are less likely to require surgery and have a higher rate of conservative treatment options [7, 29].
The referral rates from advanced physiotherapy practitioner clinic for surgery/ surgical opinion are comparable to the literature, although high variation is reported, ranging from 9 to 66% [31, 33, 35]. A reason for this variation is the difference in clinic structure, i.e. multi-profession versus single profession clinics. Referral rates for MRI were similar to those reported in previous studies (13–23% referral rate) [7, 29, 33]. We found evidence that some treatments such as dietetics was underutilised, with a very low referral rate across clinic type despite high levels of patients that were classified as overweight or obese. Similar finding has been reported previously by [6, 36].
In the current study, participants were more likely to follow an optimal care pathway if seen in orthopaedic clinic and less likely if seen in advanced physiotherapy practitioner clinic. One reason for this is that patients seen in orthopaedic more frequently had one definitive treatment at the conclusion of the specialist consultation, whereas in advanced physiotherapy practitioner clinics patients more frequently had multiple treatment outcomes. Furthermore, higher numbers of patients were given review appointments for advanced physiotherapy practitioner clinics and this may be as a result of patients trying a range of conservative treatments and therefore the outcome of these was being monitored [20]. In the future additional methods of optimising the pathway could include adopting a combined skill mix of professions, providing training for primary care clinicians and developing methods for streamlining specialist referrals to the appropriate profession [5].
Study limitations
There were missing values, especially around the BMI. Further limitations are concerned with the generalisability of findings as (1) there was a lower proportion of patients with hip pain and (2) data were collected from a single Health Board. The definition of ‘optimal pathway’ used in this study was based around efficient healthcare resource allocation in line with published guidelines and local policy/ referral guidance. The context of this study means that the application of ‘optimal and sub optimal’ pathway is subjective and will apply differently across different services. This does reflect the complexity and activity loops present within the care pathway [5]. This definition is limited as it does not take into consideration changes in the patient’s condition, patient opinion of their symptoms or characteristics and preferences of the referrers. There was inconsistency for the diagnostic procedure variable at predicting optimal care pathway. Two or more variables was associated with an optimal care pathway but one diagnostic procedure was associated with a sub-optimal care pathway. Therefore this variable needs to be interpreted with caution. Finally, it has not been established how many of those that were referred to a consultant ended up having surgery, which may have affected what was recorded as a treatment outcome, i.e. referral or surgery. Due to missing data a multivariate analysis was not conducted on KOOS/HOOS combined scores at 6 months post consultation.