Cost-utility analysis of risk-based stratified primary care for common musculoskeletal pain presentations: a cluster-randomised, controlled trial

doi:10.21203/rs.3.rs-4236308/v1

Download PDF

Research Article

Cost-utility analysis of risk-based stratified primary care for common musculoskeletal pain presentations: a cluster-randomised, controlled trial

https://doi.org/10.21203/rs.3.rs-4236308/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Risk-based stratified care (SC) has demonstrated cost-effectiveness versus usual primary care for non-specific low back pain. The STarT MSK trial investigated the cost-effectiveness of risk-based stratified care versus non-stratified usual primary care for patients with the five most common musculoskeletal pain presentations.

Methods

A cost-utility analysis was undertaken over 6-months. The base-case analysis estimated the incremental costs per additional quality-adjusted life year (QALY), using the EQ-5D-5L to generate QALYs, for the overall trial population and for each risk subgroup (low, medium, high risk) for persistent disabling pain. The base-case analysis used the intention-to-treat principle and was performed from an NHS and personal social services (PSS) perspective. Uncertainty was explored with cost-effectiveness acceptability curves. Sensitivity analyses included a healthcare and societal perspective, complete-case, and risk subgroup analyses.

Results

Risk-based stratified primary care showed similar costs to usual primary care, with a small QALY gain of 0.0041 (95% CI -0.0013, 0.0094). The incremental cost-effectiveness ratio was £1,670 per QALY with a likelihood that stratified care represents cost-effective use of resources of 73% at a willingness-to-pay threshold of £20,000 per QALY. In subgroup analyses, stratified care was only likely to be cost-effective for the subgroup of patients at high risk of poor outcome.

Conclusions

Risk-based stratified primary care for patients with the five common musculoskeletal pain presentations resulted in similar costs and small QALY gains compared to usual, non-stratified primary care overall and therefore, showed inconclusive results overall.

Trial Registration:

ISRCTN Registry ISRCTN15366334; http://www.isrctn.com/ISRCTN15366334.

economic evaluation

cost-effectiveness

cost-utility

musculoskeletal pain

stratified care

Musculoskeletal (MSK) pain conditions such as back, neck, knee, shoulder and multi-site pain are known to affect over 20 million people (almost one third of the population) in the UK [1]. Annual NHS MSK pain related costs have been estimated at £5 billion [2]. Given the prevalence and economic burden of MSK pain, there is a need to find better ways to manage MSK pain that improve patient outcomes such as pain, and make the best use of scarce healthcare resources [3, 4]. One strategy is to use a risk-based stratified care approach in primary care, whereby patient’s risk of persistent disabling pain is identified (low, medium, high risk) and treatments targeted according to these subgroups, with the aim of maximising treatment benefit and preventing unnecessary treatments and harm [5].

Risk-based stratified primary care for patients with low back pain (LBP) was evaluated in two United Kingdom (UK) studies and found to be cost-effective compared to a best conservative physiotherapy-led control and usual primary care [6, 7, 8]. A replication in a Danish setting found that stratified care resulted in lower healthcare costs, but not greater benefits in terms of cost-effectiveness [9]. In a US setting, stratified care demonstrated no significant effects on patient healthcare resource utilisation, although this was not wholly unexpected, as the stratified care intervention failed to change primary care clinician treatment decision-making [10, 11].

Despite evidence regarding risk-based stratified primary care’s cost-effectiveness for patients with LBP, it remains unclear whether a similar model may be cost-effective for a much broader group of patients consulting with MSK pain in UK primary care. The recently completed STarT MSK cluster randomised controlled trial tested the clinical effectiveness of a risk-based stratified care model for patients with back, neck, shoulder, knee and multi-site pain compared to usual primary care [12]. That trial found that stratified care did not lead to significant improvements in clinical outcomes but did change some aspects of primary care treatment decision-making for MSK patients [12]. The primary aim of this economic evaluation embedded in the STarT MSK trial was to determine the cost-effectiveness of the risk-based prognostic stratified primary care model for this broader group of patients with MSK pain compared to usual primary care.

Trial design and interventions

The main trial this paper is part of received ethical approval from the NHS REC East Midlands Nottingham 1 (Ref:16/EM/0257). The clinical trial and its findings have been reported in detail elsewhere [12]. Full details of the Keele STarT MSK risk-based stratification tool have also been published elsewhere [13, 14, 15]. Adults aged 18 years or older consulting their general practitioner for one of the five most common MSK presentations (back, neck, knee, shoulder and multi-site pain) were invited to participate in the study. The STarT MSK trial was a cluster RCT, with general practices randomised to offer the stratified care intervention or continue with usual care by a computer-generated randomisation. GPs in practices allocated to the intervention group were trained and supported to deliver stratified care via a bespoke computer-based template which housed the risk-stratification tool (the Keele STarT MSK Tool) and recommendations about a suite of matched treatment options for each risk group. The intervention aimed to improve patient outcomes (with the primary outcome being pain intensity), and to change clinician treatment decision-making to be more guideline adherent. For example, stratified care aimed to encourage GPs to prescribe less long-term opioids, neuromodulators, muscle relaxants and corticosteroid injections, send less unnecessary referrals (e.g., to imaging and specialist orthopaedics), and use less sick certification (particularly for low-risk patients); whilst encouraging more written self-management advice, simple over-the-counter analgesics, earlier referral to physiotherapy (for patients at medium or high risk), plus further GP assessment to address complexities such as comorbidities, distress, and emerging frailty (for patients at high risk). In the usual care group, patients were managed as usual without the use of formal risk-stratification tools. This study reports the economic analysis of risk-based stratified care for the overall stratified care approach and for risk-defined subgroups within a cost–utility framework.

Economic study design: overview

The within-trial cost-utility analysis estimated the cost-effectiveness of stratified care versus usual care for adults consulting in primary care with the five most common MSK pain presentations. The base case analysis was based on the intention-to-treat population and conducted from the perspective of UK National Health Service and Personal Social Services (NHS/ PSS). Costs were expressed in British pounds sterling (2019 price year) and health outcomes in quality-adjusted life-years (QALYs). The time horizon covered the period from the patient consent in the study to end of follow-up at 6 months, and all data included in this analysis were collected via self-report postal questionnaires. Costs and outcomes were not discounted as the trial was limited to 6 months follow-up. Trial findings are reported in accordance with the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) [16].

Data: measures and collection

Healthcare resource use and unit costs

The 6-month follow-up questionnaire obtained resource use data on: primary and community care contacts (face-to-face general practitioner, practice nurse, community therapy, and other primary contacts), hospital based services (e.g. consultants, outpatient appointments, physiotherapy, inpatient admissions, diagnostic tests, scans and surgical procedures), private costs incurred by patients (including private physiotherapy and other private medical use), and time off work and reduced work performance (presenteeism) [17] related to their MSK pain problem. Prescribed medication was obtained from an extraction and review of primary care medical records at the participating general practices.

Unit costs were obtained from several standard sources (e.g. from medical record prescription databases and the British National Formulary, the PSSRU Unit Costs of Health and Social Care compendium, NHS Reference costs) and are reported in Additional file 1 in 2019 prices [18, 19, 20]. Implementation and intervention costs including the cost of GP training sessions, and support to use the electronic template were not included in the analysis.

Work-related outcomes

To estimate productivity costs, self-reported details of participant employment status including days off work and reduced productivity while at work (presenteeism) were collected at baseline and 6-months. Presenteeism was based on the rate at which MSK pain had affected the participants’ work ability while at work (presenteeism) on a validated 0–10 single-item presenteeism scale [17]. Presenteeism time loss was generated by translating the self-reported scores obtained into work loss based on the actual number of days worked during this period [21]. Productivity costs were estimated by multiplying productivity loss outputs with the average wage rate identified from annual earnings data [22]. The analyses used the human capital approach [21].

Measurement of outcomes

Health-related quality of life was assessed in the initial patient questionnaire (shortly after their GP consultation) and again at 6-month follow-up using the Euroqol’s EQ-5D-5L questionnaire. In line with current guidelines, responses were converted into index scores using the interim crosswalk value set for mapping from the EQ-5D-5L to the EQ-5D-3L (23). Quality-adjusted life-years (QALYs) were generated for each patient using the area under the baseline-adjusted utility curve, assuming linear interpolation between two follow-up time-points [24]. EQ-5D-5L values and QALYs over 6 months were then reported by treatment group and presented as means and standard deviations.

Statistical Analysis

The statistical analysis was conducted on an intention-to-treat basis, and in accordance with current cluster RCT guidelines [25, 26]. A multi-level modelling statistical approach taking into consideration clustering in cost and effect data was adopted. Missing EQ-5D-5L and total cost data were imputed using multiple imputation techniques [27] to address issues with incomplete data. The imputed dataset informed the base-case and all subgroup and sensitivity analyses, except for the complete-case analysis. The imputation included 25 imputed datasets, and Rubin’s rule was used to combine the imputed datasets into one final imputed variable [28]. The cost-effectiveness analysis assessed the joint estimation of costs and incremental QALYs (SC value minus UC value) to estimate the incremental cost per QALY (an estimate of the amount required to achieve an additional QALY) within each subgroup. In estimating incremental QALYs, between group differences were compared using multilevel regression modelling techniques. QALYs were adjusted for clustering, EQ-5D-5L score, age and gender variables in the multivariate regression model using data from the initial questionnaire. Resource use and cost summary statistics were generated by treatment group over 6 months follow-up. Total healthcare costs over the study period were calculated by multiplying the resource items used by the respective unit cost and summing over all items. Between group differences were compared using generalised linear models adjusting for clustering. Bootstrapped 95% confidence intervals for the between-group differences in cost estimates were also reported. Statistical analysis was performed using Stata V.16 [29].

Cost-effectiveness analysis

Separate generalised equation models, controlling for clustering were used to estimate the mean incremental costs and QALYs for stratified care relative to usual care. Uncertainty around these estimates was estimated using cost-effectiveness acceptability curves (CEACs), which link the probability of SC being cost-effective to a range of potential threshold values (λ) that the healthcare system may be willing to pay for an additional QALY gained. CEACs were estimated using a Net Monetary Benefit (NMB) regression approach [25]. Net Monetary Benefit (NMB) was defined as λ x (Δ effect_i) - Δ cost_i, where Δ effect_i is the incremental person-level outcome associated with the stratified care intervention, and (Δ cost_i), the additional costs due to stratified care, and λ = willingness to pay per unit of outcome gain. Using the output, we plotted CEACs, to provide a graphical display of the probability that stratified care is cost-effective across a wide range of cost-effectiveness thresholds. Results for probability of cost-effectiveness were reported for the £20,000 per QALY threshold, currently used by NICE to determine cost-effectiveness of interventions for the NHS.

Sensitivity analyses

Sensitivity analyses were conducted to investigate variations in methodological and data collection approaches, and these included: i) A complete-case analysis which required valid EQ-5D-5L responses in the initial patient questionnaire and at 6-month follow-up and complete resource use data at 6-month follow-up; ii) broadening the perspective of the analysis to healthcare and societal perspectives by capturing private costs and wider societal costs and their impact on relative cost-effectiveness of the interventions; and iii) additional exploratory subgroup analysis to investigate the cost-effectiveness of stratified care compared with usual care for participants in each risk subgroup (low, medium and high risk).

Response rates and data completion

Estimates for the base-case analysis were based on all 1211 participants following multiple imputation. Complete cost estimates were available for 978 (81%) responders at 6 months. For the EQ-5D-5L, complete outcome data were available for 1070 (88%) participants including minimum data collection at 6 months. After excluding missing EQ-5D-5L items, 1019 (84%) participants had a valid EQ-5D-5L score at 6 months.

Health-related quality of life outcomes

Mean EQ-5D-5L scores from the initial questionnaire and at 6-months and mean QALYs are reported in Table 1. There was overall improvement over the 6 months reflected by increased EQ-5D-5L scores from baseline through to 6 months in both trial groups. The adjusted and imputed mean QALYs over 6 months were higher for SC than for the UC group but differences between the groups were very small 0.0041 (95% CI -0.0013, 0.0095). The result did not change with the unadjusted or complete-case analyses.

Table 1

Descriptive and incremental health outcomes over 6 months
Health outcomes	Stratified care	Usual care	Mean difference^b (95% CI)
Primary (Imputed) analysis^a	n = 534	n = 677
Baseline EQ-5D-5L	0.5572 (0.2293)	0.5542 (0.2349)	0.0011 (-0.0453, 0.0476)
6-month EQ-5D-5L	0.6715 (0.2192)	0.6512 (0.2308)	0.0166 (-0.0233, 0.0566)
Unadjusted total QALYs	0.3072 (0.1012)	0.3014 (0.1042)	0.0054 (-0.0134, 0.0242)
Adjusted total QALYs^c	0.3063	0.3021	0.0041 (-0.0013, 0.0095)
Complete-case analysis	n = 445	n = 574
Unadjusted total QALYs	0.3095 (0.1034)	0.3032 (0.1063)	0.0068 (-0.0141, 0.0277)
Adjusted total QALYs^c	0.3080	0.3027	0.0053 (-0.0005, 0.0112)
QALYs = quality-adjusted-life-years. ^aBase-case imputed dataset ^bDifference=Stratified Care-Usual care by generalized linear latent and mixed (gllamm) models adjusting for clustering. Reported CIs were generated using generalised linear regression methods. ^cAdjusted for baseline utility.

Resource use and cost

Table 2 reports the disaggregated healthcare resource use over 6 months. Corresponding disaggregated MSK pain-related costs over 6 months are reported in Table 3. Over the 6 months, there were minimal differences overall in the reported use of primary and secondary healthcare services between the two trial groups, except for the number of visits to NHS and private physiotherapy and NHS scans and injections. The number of NHS physiotherapy visits was higher in the stratified care group, while the number of private physiotherapy visits, NHS scans (4.8% compared with 3.1%) and injections (4.3% compared with 1%) were higher in the usual care group (See Table 3).

Table 2

**Healthcare resource use per patient, by treatment group, over 6-months**
Healthcare resource	Stratified care	Usual care
Primary care contacts:	n = 415	n = 563
Primary care general practitioner	0.96 (1.60)	1.15 (2.11)
Primary care nurse	0.08 (0.57)	0.13 (0.65)
Primary care physiotherapist	0.35 (1.67)	0.29 (1.17)
Hospital-based care:
NHS consultant	0.33 (1.32)	0.37 (0.99)
NHS other physiotherapist	1.64 (2.64)	0.71 (1.87)
NHS acupuncturist	0.06 (0.59)	0.03 (0.44)
NHS osteopath	0.002 (0.05)	0.02 (0.22)
Private consultant	0.02 (0.23)	0.06 (0.40)
Private physiotherapist	0.21 (1.24)	0.37 (1.99)
Private acupuncturist	0.06 (0.55)	0.06 (0.42)
Private osteopath	0.29 (1.28)	0.26 (1.57)
Other Hospital-based care (n, %)
NHS surgery	10 (2.2%)	11 (2.0%)
Shoulder	2	1
Neck	0	1
Spine/Back	1	1
Hip	1	4
Knee	6	4
Private surgery	1 (0.2%)	2 (0.4%)
Back	1	0
Knee	0	1
Shoulder	0	1
NHS Scans	13 (3.1%)	27 (4.8%)
NHS MRI investigations	64 (15.4%)	67 (11.9%)
NHS Blood tests	1 (0.2%)	5 (0.9%)
NHS injections	4 (1.0%)	24 (4.3%)
Private Scans	1 (0.2%)	1 (0.2%)
Private MRI investigation	12 (2.9%)	13 (2.3%)
Private Injections	1 (0.2%)	0
SC = Stratified Care; UC = Usual Care, NHS, National Health Service

Table 3

**Healthcare costs (£) per patient, by treatment group**
	Cost (£)
Healthcare resource	Stratified Care	Usual Care
Primary care contacts:	n = 415	n = 563
Primary care general practitioner	31.81 (52.81)	37.87 (69.58)
Primary care nurse	3.14 (23.77)	5.52 (27.11)
Primary care physiotherapist	20.40 (96.94)	17.10 (68.11)
Hospital-based care
NHS consultant	51.72 (209.25)	59.31 (158.67)
NHS other physiotherapist	95.18 (152.96)	41.00 (108.72)
NHS Acupuncturist	3.49 (34.15)	1.95 (25.70)
NHS Osteopath	0.14 (2.85)	0.93 (12.67)
NHS other professionals	3.54 (41.36)	5.25 (78.07)
Private consultant	3.83 (36.45)	9.32 (64.00)
Private Physiotherapist	12.44 (72.11)	21.43 (115.57)
Private Acupuncturist	3.63 (31.79)	3.50 (24.21)
Private Osteopath	16.63 (74.38)	15.25 (91.05)
NHS private other	2.65 (33.98)	2.37 (32.27)
Other hospital-based care
NHS surgery^†	110.07 (728.99)	106.34 (808.99)
NHS investigations and treatments^{† †}	31.61 (86.22)	61.13 (254.14)
Private surgery^†	2.80 (57.04)	16.38 (284.23)
Private investigations/treatments^{† †}	6.86 (63.76)	3.93 (31.63)
Estimates for the base-case (imputed) analysis
Total NHS cost	356.36 (864.01)	343.44 (942.92)
Mean difference (95% CI)^#	6.85 (-107.82, 121.54)
^# Between-group difference in mean scores (SC– UC) by generalized linear latent and mixed (gllamm) models adjusting for clustering, ^†Includes shoulder, neck, spine, back, hip, knee injury; ^†† Includes investigations such as MRI scans, CT scans and ultrasound scans, and injections.

Estimation of Cost-utility

Table 4 provides a summary of the incremental costs, outcomes, net monetary benefit results and cost-utility analysis using data from the base-case NHS/PSS perspective. Point estimates of incremental costs and incremental QALYs were similar between the two groups, but SC was associated with health gains of 0.0041 QALYs (95% CI -0.0013, 0.0094). The net monetary benefit was £132 and the resulting ICER was £1,670 per QALY gained with a 73% chance of SC being cost-effective at a cost-effectiveness threshold of £20,000 per QALY gained. The CEAC shows the probability that the SC intervention is potentially cost-effective at different levels of willingness-to-pay for a QALY (Fig. 1).

Table 4

Cost-utility analysis outcomes over 6 months follow-up
					Probability SC is cost-effective at cost-effectiveness threshold of:
	Mean incremental costs (95% CI), £	Mean incremental QALYs (95% CI)		ICER (£ per QALY)	£20,000 per QALY	£30,000 per QALY	£50,000 per QALY
Base-case analysis	6.85 (-107.82, 121.54)	0.0041 (-0.0013, 0.0094)		£1,670^†	0.73	0.73	0.73
Sensitivity analysis 1: Alternative-perspectives
Healthcare perspective	-5.92 (-116.32, 104.46)		0.0041 (-0.0013, 0.0094)	SC Dominant	0.75	0.75	0.74
Societal perspective	65.01 (-195.56, 325.59)		0.0041 (-0.0013, 0.0094)	£15,856^†	0.62	0.61	0.63
Sensitivity analysis 2: Complete-case analysis††
NHS cost (£) Mean (SD)	11.61 (-118.84, 142.08)	0.0053 (-0.0005, 0.0112)		£2,190^†	0.75	0.74	0.73
Healthcare cost (£) Mean (SD)	-10.21 (-139.95, 119.53)	0.0053 (-0.0005, 0.0112)		SC Dominant	0.78	0.77	0.75
Societal cost (£) Mean (SD)	19.05 (-295.94, 334.04)	0.0053 (-0.0005, 0.0112)		£3,594^†	0.74	0.72	0.71
The between-group mean differences in QALYs adjusted for age, sex, treatment allocation, baseline health-related quality of life. ^†Mean ICERs in north east quadrant of the cost-effectiveness plane where SC is more costly and more effective. NHS, National Health Service, QALY, Quality-Adjusted Life Year. ^* Mean differences adjusted for clustering in cost and QALY outcomes. ^†† Complete-case dataset. ICER = Incremental cost-effectiveness ratio; CI = confidence interval.

Figure 1 - Cost Effectiveness Acceptability Curve (Base-case)

Work-related outcomes

Less than half the patients in the trial were in paid employment at the 6-month follow-up; 44.1% in the SC group and 39.4% in the UC group. Similar rates were observed at the start of the trial. Data on employment and time-off-work are reported in Table 5. Of those who at 6-months reported being in employment, 47 (23.6%) patients in the SC group reported time off paid work (employment) because of MSK pain, compared with 57 (23.2%) in the UC group.

Table 5

**Description of work-related outcomes for participants in paid employment by treatment group.**
Work-related outcomes	Stratified care	Usual care
Baseline: working in paid employment (n, %)	275 (53.8)	286 (43.3)
Baseline: reported time off work during the last 6 months (n, %)	91 (30.6)	101 (30.2)
Working in paid employment at 6-months (n, %)	177 (44.1)	219 (39.8)
Performance at work at 6 months mean, SD) ^b	3.41 (2.71)	3.57 (2.89)
Reported time off work at 6-months (n, %)	47 (23.6)	57 (23.2)
Days off-work during the last 6 months (mean, SD)^a	6.13 (20.05)	6.48 (19.66)
Mean difference^c (95% CI)	-1.09 (-5.87, 3.68)
Productivity costs during the last 6 months (mean, SD)^a	698.81 (2285.24)	738.14 (2240.95)
Mean difference^d (95% CI)	-124.89 (-669.73, 419.93)
^aThe evaluation of work-related outcomes and the estimation of indirect costs focussed on the subsample of respondents in paid employment at 6 months (396/1211). ^bMean performance at work on a scale of 0 to 10 where 0 indicates work performance not affected. ^c Mean difference of reported days off-work over 6 months and adjusted for paid employment at baseline and clustering. ^dProductivity costs obtained from days off-work over 6 months and adjusted for paid employment at baseline and clustering.

During the 6-month follow-up period, the mean number of days off work were similar between the groups (SC group, 6.1 days and the UC group, 6.5 days), with costs translating into £698 for SC and £738 for UC. Overall, reduced productivity (presenteeism) was very similar between the two groups (mean scores of 3.41 and 3.57 for SC and UC respectively).

Sensitivity analyses

Similar results were replicated in all sensitivity analyses that explored variations in perspective including healthcare and societal perspectives, that is, costs were very similar, but SC was associated with marginal health gains. The probability that SC was cost-effective at £20,000 per QALY was 76% and 61% from a healthcare and societal perspective, respectively. Within the complete-case analysis, the stratified care intervention was associated with a 75% probability of cost-effectiveness, overall, for the UK primary care setting. The net monetary benefit value was £162 (λ=£20,000); and the ICER was £2,190 per QALY gained. Details of the sensitivity analyses are reported in Table 5 and Additional file 2.

Subgroup analysis

Results in the sub-group analysis showed there was overall improvement in quality of life over the 6 months reflected by increased EQ-5D-5L scores from the initial patient questionnaire through to 6 months in both trial groups for all risk subgroups. SC was dominated in the low risk and medium risk subgroups, i.e., stratified care was slightly more costly and less effective than usual care. For those at high risk, SC dominated UC over 6 months follow-up, that is, greater health benefits (0.0129 additional QALYs) and slightly reduced NHS costs (-£65.96). This was in part driven by the higher number of spinal injections and spinal imaging in the UC group. Details are provided in Additional file 3 and 4.

This is the first study to estimate the cost-effectiveness of risk-based stratified primary care for a broader group of patients with common MSK pain presentations, compared to usual primary care. The results provide the first detailed overall and risk-subgroup specific cost-effectiveness findings. Overall, the costs of care were similar, but risk-based stratified care produced a small incremental gain in quality of life compared with usual care. The small differences could be explained by, for example, the higher NHS physiotherapy visits (costing £54 more) in the SC group balanced out by more NHS imaging and more spinal injections (costing £30 more on average) in the UC group. For both costs and QALYs, confidence intervals were wide and highlighted the uncertainty in the cost-effectiveness estimates. Whilst the findings could suggest that risk-based stratified care potentially represents cost-effective use of healthcare resources, the very small differences suggest caution in that interpretation. Furthermore, subgroup analyses suggested that SC was likely to be cost-effective, for patients at high risk of persistent disabling pain, but the sample sizes were quite small for any meaningful interpretation.

Comparison with other literature

There are currently no studies that have investigated risk-based stratified care for a much broader group of patients with MSK pain. However, stratified primary care using a different tool (the STarT Back tool) and three matched treatments (one for each risk subgroup) for patients with LBP has previously been found to be a cost-effective treatment strategy compared with current best practice and compared with usual primary care [7, 8]. In the LBP studies, the stratified care model was found to improve health related quality of life (QALYs) and reduce overall healthcare resource use and days off work overall and across all risk subgroups although cost-effectiveness was greatest in the high-risk group. A broad replication in a Danish primary care context found that stratified care resulted in lower healthcare costs, but did not result in greater benefits in terms of cost-effectiveness [9]. More recently, a different stratified care model combining risk of poor outcome with clinical indicators of severity for patients with sciatica found no evidence that this was a more cost-effective use of healthcare resources when compared with non-stratified usual care [30]. In comparison, our findings show that the risk-based stratified care model tested here could potentially be cost-effective within the context of the large volume of patients consulting in primary care with the five most common MSK pain presentations and specifically this was driven primarily by benefits for patients who are at high risk of persistent disabling pain.

Strengths and limitations

A strength of this study is the large sample (n = 1211) with comprehensive resource use information reported both from an NHS and societal perspective including wider societal costs of work absence. The analysis adopted robust approaches and presents disaggregated results by risk subgroups. However, there were some limitations. Resource use and work-related data were primarily collected using a self-report approach and supplemented by medical records review for prescription data. A limitation with self-report is that respondents can potentially under-report utilisation, particularly over longer periods of recall [31]. However, the period of recall here was only 6 months and any under-reporting would likely be the same across both groups. Nevertheless, the method provides a pragmatic and efficient approach to data collection in the absence of routine data sources without over-burdening patients and researchers and has been used extensively in trial-based evaluations [32]. Also, in this study whilst GPs completed a bespoke electronic template regarding the treatments they offered to participants in the stratified care group of the trial, treatment decisions and referrals to other services do not necessarily translate into actual attended appointments. Additionally, EQ-5D-5L data were only collected in the initial patient questionnaire and at 6-month follow-up; therefore, changes in quality of life related to interventions at intermediate time-points could not be incorporated into the QALY calculation. An exploratory subgroup analysis is provided, although the trial was not powered to detect differences at the level of each risk subgroup, and therefore robust conclusions are only applicable to the overall comparison of stratified care versus usual primary care. A further limitation is that intervention implementation costs were not included in the analysis. However, given that these costs are often small when calculated per patient and the minimal differences in costs and minimal health gains in favour of stratified care, the inclusion of this cost would not be expected to impact the overall results.

The overall cost-effectiveness of this model of stratified primary care in the UK for patients with the five most common MSK pain presentations must be considered within the context of the clinical findings [12]. The STarT MSK trial found no significant differences in patient outcomes from stratified care (combining risk stratification with matched treatment options) compared to usual care, but reported benefits in some aspects of GP clinical decision-making for example in prescribing more over-the-counter analgesics and greater provision of written information to patients. However, in contrast to the clinical outcomes, considerations need to be made for the fundamentally different research questions and analytical paradigms of clinical and economic evaluation research with less consideration on significance levels [33]. The health economics results are different in terms of effectiveness outcomes between stratified care and usual care based on total QALYs but similar in terms of the exploratory subgroup analyses that showed potential benefits for patients at high risk of persistent disabling pain. However, it is important to note that the difference in effectiveness was small and not statistically significant and the QALYs were interpreted primarily in a comparative sense. This is in line with standard health economics cost-effectiveness methodology that estimates costs and effects jointly [34].

In conclusion, this economic analysis demonstrated that the risk-based stratified care for UK primary care patients with the five most common musculoskeletal pain presentations did not lead to superior clinical outcomes, and offers minimal benefits in health economics outcomes.

CEAC

cost-effectiveness acceptability curves

CHEERS

Consolidated health economic evaluating reporting standard

EQ-5D

EuroQoL five dimension

HR-QoL

Health related quality of life

ICER

Incremental cost-effectiveness ratio

LBP

Low back pain

MSK

Musculoskeletal

NHS/ PSS

National Health Service and Personal Social Services (NHS/ PSS)

NMB

Net Monetary Benefit

Osteoarthritis

QALY

Quality adjusted life years

Stratified care

Ethics approval and consent to participate:

The main trial this paper is part of received ethical approval from the NHS REC East Midlands Nottingham 1 (Ref:16/EM/0257). All participants provided written informed consent to participate in the research. All methods reported were performed in accordance with the relevant guidelines and regulations as outlined in the Declaration of Helsinki (2013).

Consent for publication

N/A

Availability of data and materials

Metadata, including the study protocol, data dictionaries, and key study documents (patient information leaflet, blank or coded case report forms, and consent form), will be deposited on a publicly accessible repository. De-identified individual participant data that underlie the results from this trial will be securely stored on servers approved by a government-backed cyber security scheme and made available to bona-fide researchers upon reasonable request via our controlled access procedures. Unless there are exceptional circumstances, data will be available upon publication of the main trial findings and with no end date. Data requests and enquiries should be directed to [email protected]. We encourage collaboration with those who collected the data, to recognise and credit their contributions. The data generated from this trial will remain the responsibility of the study sponsor. Release of data will be subject to a data use agreement between the sponsor and the third party requesting the data. De-identified individual patient data will be encrypted upon transfer.

Competing interests

The authors have no conflict of interest to declare.

Funding

This article presents independent research work funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research scheme (grant number: RP-PG-1211-20010); and Versus Arthritis through a Centre of Excellence grant [grant reference: 20202). Nadine Foster is a NIHR Senior Investigator and was supported through an NIHR Research Professorship (NIHRRP-011-015). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Authors' contributions

JCH, AC, ML, KMD, and NEF conceived the trial. NEF, JCH, ML, and KMD obtained the study funding. NEF was the chief investigator. All authors participated in the design and conduct of the trial. JK produced the first draft of the manuscript. KB, ML, SW, JK, and SJ led the statistical and health economics analysis with input from JCH and NEF. All authors contributed to the drafting and approval of the final manuscript. JK had final responsibility for the decision to submit for publication. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Acknowledgements

The views expressed in this publication are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. Funders of the study had no role in the design of the study, data collection, analysis and interpretation of data, or in writing the manuscript. We thank the STarT MSK Programme Steering Group (PSC) that oversaw the research, the staff from Keele Clinical Trials Unit, in particular Steff Garvin, Andrea Cherrington and Susie Hennings, the Patient and Public Involvement and Engagement group for their support and input, particularly John Murphy and Angie Emery, the wider STarT MSK programme research team (Benjamin Saunders, Vince Cooper, Hollie Birkinshaw, Elaine Hay, Danielle van der Windt, Christian Mallen, Tom Sanders, Bernadette Bartlam, Ebenezer Afolabi, Ying Chen, Carolyn Chew-Graham, Chris Main, and Rhian Hughes), and the patients and clinicians for their participation.

Author details

a) Health Economics Unit, Institute of Applied Health Research, University of Birmingham, Birmingham, UK

b) Keele Clinical Trials Unit (CTU), David Weatherall Building, Keele University, Staffordshire, ST5 5BG

c) Primary Care Centre Versus Arthritis, School of Medicine, Keele University, Staffordshire, ST5 5BG, UK

d) STARS Education and Research Alliance, Surgical Treatment and Rehabilitation Service, The University of Queensland and Metro North Health, Queensland, Australia

Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2020 (GBD 2019) Results. Institute for Health Metrics and Evaluation (IHME), Seattle.
NHS England (2015). CCG Programme Budgeting Benchmarking Tool 2013/14.
Foster NE, Hartvigsen J, Croft PR. Taking responsibility for the early assessment and treatment of patients with musculoskeletal pain: a review and critical analysis. Arthritis Research Therapy. 2012;14(1):205.
Foster NE, Hill JC, O'Sullivan P, Hancock M. Stratified models of care. Best Practice & Research Clinical Rheumatology. 2013;27(5):649-61.
Hingorani AD, Windt DA, Riley RD, Abrams K, Moons KG, Steyerberg EW et al; PROGRESS Group (PROGRESS) 4: stratified medicine research. BMJ.;346: e5793.
Hill JC, Whitehurst DG, Lewis M, Bryan S, Dunn KM, Foster NE et al. Comparison of stratified primary care management for low back pain with current best practice (STarT Back): a randomised controlled trial. Lancet. 2011;378(9802):1560-71.
Foster NE, Mullis R, Hill JC, Lewis M, Whitehurst DG, Doyle C et al; IMPaCT Back Study team. Effect of Stratified Care for Low Back Pain in Family Practice (IMPaCT Back): A Prospective Population-Based Sequential Comparison. Annals of Family Medicine. 2014;12(2):102-11
Hall J, Jowett S, Lewis M, Oppong R (2020). The STarT Back stratified care model for non-specific low back pain: a model-based evaluation of long-term cost-effectiveness. Pain. 2021;162(3):702-710.
Morsø L, Olsen Rose K, Schiøttz-Christensen B, Sowden G, Søndergaard J, Christiansen DH. Effectiveness of stratified treatment for back pain in Danish primary care: A randomized controlled trial. European Journal Pain. 2021 Oct;25(9):2020-2038. doi: 10.1002/ejp.1818. Epub 2021 Jul 2. PMID: 34101953; PMCID: PMC8518659.
Cherkin, D., Balderson, B., Wellman, R., Hsu C, Sherman KJ, Evers SC et al. Effect of Low Back Pain Risk-Stratification Strategy on Patient Outcomes and Care Processes: the MATCH Randomized Trial in Primary Care. Journal of General Internal Medicine .33, 1324–1336 (2018). https://doi.org/10.1007/s11606-018-4468-9
Delitto A, Patterson CG, Stevans JM, Freburger JK, Khoja SS, Schneider MJ et al. Stratified care to prevent chronic low back pain in high-risk patients: The TARGET trial. A multi-site pragmatic cluster randomized trial. EClinicalMedicine. 2021 Mar 30; 34:100795. doi: 10.1016/j.eclinm.2021.100795. PMID: 33870150; PMCID: PMC8040279.
Hill JC, Garvin S, Bromley K, Saunders, B, Kigozi J, Cooper V et al (2022). Risk-based stratified primary care for common musculoskeletal pain presentations (STarT MSK): a cluster-randomised, controlled trial. The Lancet Rheumatology, Volume 0, Issue 0.
Protheroe J, Saunders B, Bartlam B, Kate M, Dunn KM, Cooper V et al. Agreeing risk stratified matched treatment options for musculoskeletal pain in primary care: a consensus group study. BMC Musculoskeletal Disorder. 2019; 20:271.
Protheroe J, Saunders B, Hill JC, Chudyk A, Foster NE, Bartlam B et al. Testing Stratified Care for Musculoskeletal Pain in General Practice: Integrating Clinician Support and Intervention Design. Research Square; 2021. DOI: 10.21203/rs.3.rs-209375/v1.
Dunn KM, Campbell P, Lewis M, Hill JC, van der Windt DA, Afolabi E et al. Refinement and validation of a tool for stratifying patients with musculoskeletal pain. European Journal Pain. 2021 Nov;25(10):2081-2093. doi: 10.1002/ejp.1821. Epub 2021 Jul 3. PMID: 34101299.
Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, Augustovski F, Briggs AH, Mauskopf J, Loder E; CHEERS Task Force. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. Value Health. 2013 Mar-Apr;16(2):e1-5. doi: 10.1016/j.jval.2013.02.010. PMID: 23538200.
Kigozi J, Lewis M, Jowett S, Barton P, Coast J. (2014). Construct Validity and Responsiveness of the single-item presenteeism question in patients with lower back pain for the measurement of presenteeism. Spine 2014;39:409–16.
BNF, 2019. Joint Formulary Committee. BNF - British National Formulary. London: BMJ Group and Pharmaceutical press; 2019.
Curtis and Burns, 2019. Curtis L, Burns A. PSSRU: Unit Costs of Health & Social Care. University of Kent, Canterbury, United Kingdom: Personal Social Service Research Unit (PSSRU); 2019.
Department of Health. National Schedule of Reference Costs: 2018–2019. London: Department of Health; 2019.
Krol M, Brouwer W. How to Estimate Productivity Costs in Economic Evaluations. PharmacoEconomics 2014;4:335–44.
Office for National Statistics. Annual survey of hours and earnings. n.d. URL: http://www.ons.gov.uk. 2017. (Accessed 1 March 2019).
van Hout B, Janssen M, Feng V, Kohlmann T, Busschbach J, Golicki D et al. Interim scoring for the EQ-5D-5L: mapping the EQ-5D-5L to EQ-5D-3L value sets. Value Health 2012; 15:708–15.
Manca A, Hawkins N, Sculpher M. Estimating mean QALYs in trial-based cost-effectiveness analysis: The importance of controlling for baseline utility Article in Health Economics. Health Economics 2005;14:487–96.
Gomes M, Ng ESW, Grieve R, Nixon R, Carpenter J, and Thompson SG: Developing appropriate methods for cost-effectiveness analysis of cluster randomized trials. Med Decis Making 2012, 32:350-361.
Ng ES, Diaz-Ordaz K, Grieve R, Nixon RM, Thompson SG, Carpenter JR: Multilevel models for cost-effectiveness analyses that use cluster randomised trial data: an approach to model choice.
Schafer J. Multiple imputation: a primer. Stat Methods Med Res 1999; 8:3–15.
Rubin D, Schenker N. Multiple imputation in health‐are databases: An overview and some applications. Stat Med 1991;10:585–98.
StataCorp. 2019. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC.
Konstantinou, K., Lewis, M., Dunn, K. M., Ogollah, R., Artus, M., Hill, et al (2020). Stratified care versus usual care for management of patients presenting with sciatica in primary care (SCOPiC): a randomised controlled trial. The Lancet Rheumatology;2020; 2(7), e401–e411. https://doi.org/10.1016/S2665-9913(20)30099-0.
Petrou S, Murray L, Cooper P, and Davidson LL. The accuracy of self-reported healthcare resource utilization in health economic studies. Int J Technol Assess Health Care 2002; 18:705–10.
Whitehurst DG, Bryan S, Lewis M, Hay EM, Mullis R, Foster NE. Implementing stratified primary care management for low back pain: cost-utility analysis alongside a prospective, population-based, sequential comparison study. Spine (Phila Pa 1976). 2015 Mar 15;40(6):405-14. doi: 10.1097/BRS.0000000000000770. PMID: 25599287.
Whitehurst, D.G., Bryan, S. Trial-based clinical and economic analyses: the unhelpful quest for conformity. Trials 14, 421 (2013). https://doi.org/10.1186/1745-6215-14-421.
Dakin H, Wordsworth S. Cost-minimisation analysis versus cost-effectiveness analysis, revisited. Health Econ 2013; 22:22–34.

No competing interests reported.

Additionalfile.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Cost-utility analysis of risk-based stratified primary care for common musculoskeletal pain presentations: a cluster-randomised, controlled trial

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Trial Registration:

Figures

Introduction

Patients and Methods

Trial design and interventions

Economic study design: overview

Data: measures and collection

Healthcare resource use and unit costs

Work-related outcomes

Measurement of outcomes

Statistical Analysis

Cost-effectiveness analysis

Sensitivity analyses

Results

Response rates and data completion

Health-related quality of life outcomes

Resource use and cost

Estimation of Cost-utility

Work-related outcomes

Sensitivity analyses

Subgroup analysis

Discussion

Comparison with other literature

Strengths and limitations

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1