The Impact of Regression to the Mean on Economic Evaluation in Quasi-Experimental Pre-Post Studies: The Example of Total Knee Replacement Using Data from the Osteoarthritis Initiative

Home Telemonitoring Technology for Patients With Heart Failure: Cost-Consequence Analysis of a Pilot Study

Background

Heart failure (HF) is a costly health condition and a major public health problem. It is estimated that 2%-3% of the population in developed countries has HF, and the prevalence increases to 8% among patients aged ≥75 years. Home telemonitoring is a form of noninvasive, remote patient monitoring that aims to improve the care and management of patients with chronic HF. Telehealth for Emergency-Community Continuity of Care Connectivity via Home-Telemonitoring (TEC4Home) is a project that implements and evaluates a comprehensive home monitoring protocol designed to support patients with HF as they transition from the emergency department to home.

Objective

The aim of this study is to assess the cost of using the home monitoring platform (TEC4Home) relative to usual care for patients with HF.

Methods

This study is a cost-consequence analysis of the TEC4Home pilot study. The analysis was conducted from a partial societal perspective, including direct and indirect health care costs. The aim is to assess the costs of the home monitoring platform relative to usual care and track costs related to health care utilization during the 90-day postdischarge period.

Results

Economic analysis of the TEC4Home pilot study showed a positive trend in cost savings for patients using TEC4Home. From both the health system perspective (Pre TEC4Home cost per patient: CAD $2924 vs post TEC4Home cost per patient: CAD $1293; P=.01) and partial societal perspective (Pre TEC4Home cost per patient: CAD $2411 vs post TEC4Home cost per patient: CAD $1108; P=.01), we observed a statistically significant cost saving per patient.

Conclusions

In line with the advantages of conducting an economic analysis alongside a feasibility study, the economic analysis of the TEC4Home pilot study facilitated the piloting of patient questionnaires and informed the methodology for a full clinical trial.

Healthcare Funding Decisions and Real-World Benefits: Reducing Bias by Matching Untreated Patients

Governments and health insurers often make funding decisions based on health gains from randomised controlled trials. These decisions are inherently uncertain because health gains in trials may not translate to practice owing to differences in the population, treatment use and setting. Post-market analysis of real-world data can provide additional evidence but estimates from standard matching methods may be biased when unobserved characteristics explain whether a patient is treated and their outcomes. We propose a new untreated matching approach that can reduce this bias. Our approach utilises the outcomes of contemporaneous untreated patients to improve the matching of treated and historical control patients. We assess the performance of this new approach compared to standard matching using a simulation study and demonstrate the steps required using a funding decision for prostate cancer treatments in Australia. Our simulation study shows that our new matching approach eliminates nearly all bias when unobserved treatment selection is related to outcomes, and outperforms standard matching in most scenarios. In our empirical example, standard matching overestimated survival by 15% (95% confidence interval 2-34) compared to our untreated matching approach. The health gains estimated using our approach were slightly lower than expected based on the trial evidence, but we also found evidence that in practice prescribers ceased prior therapies earlier, treated a more vulnerable population and continued treatment for longer. Our untreated matching approach offers researchers a new tool for reducing uncertainty in healthcare funding decisions using real-world data.

Total Hip and Knee Arthroplasties Are Highly Cost-Effective Procedures: The Importance of Duration of Follow-Up

BACKGROUND

Total hip and knee arthroplasties (THA/TKA) are clinically effective but high cost procedures. The aim of this study is to perform a cost-effectiveness analysis of THA and TKA in the New Zealand (NZ) healthcare system.

METHODS

Data were collected from 713 patients undergoing THA and 520 patients undergoing TKA at our local public hospital. SF-6D utility values were obtained from participants preoperatively and 1-year postoperatively, and deaths and any revision surgeries from patient records and the New Zealand Joint Registry at minimum 8-year follow-up. A continuous-time state-transition simulation model was used to estimate costs and health gains to 15 years. Quality-adjusted life years (QALYs), treatment costs, and incremental cost-effectiveness ratios (ICERs) were calculated to determine cost effectiveness. ICERs below NZ gross domestic product (GDP; NZ$60 600) and 0.5 times GDP per capita were considered "cost effective" and "highly cost effective" respectively.

RESULTS

Cumulative health gains were 2.8 QALYs (THA) and 2.3 QALYs (TKA) over 15 years. Cost effectiveness improved from ICERs of NZ$74,400 (THA) and NZ$93,000 (TKA) at 1 year to NZ$6000 (THA) and NZ$7500 (TKA) at 15 years. THA and TKA were cost effective after 2 years and highly cost effective after 3 years. QALY gains and cost effectiveness were greater in patients with worse preoperative functional status and younger age.

CONCLUSION

THA and TKA are highly cost-effective procedures over longer term horizons. Although preoperative status and age were associated with cost effectiveness, both THA and TKA remained cost effective in patients with less severe preoperative scores and older ages.

Patient-reported outcome measures (PROMs): can they be used to guide patient-centered care and optimize outcomes in total knee replacement?

PURPOSE

As patient-reported outcome measures (PROMs) are increasingly integrated into clinical practice, there is a need to translate collected data into valuable information to guide and improve the quality and value of patient care. The purpose of this study was to investigate health-related quality-of-life (QoL) trajectories in the 5 years following total knee replacement (TKR) and the patient characteristics associated with these trajectories. The feasibility of translating QoL trajectories into valuable information for guiding patient-centered care was also explored.

METHODS

Data on patients who underwent TKR between 2006 and 2011 from a single-institution registry were extracted including patient-reported QoL (captured using the Short Form Survey (SF-12) instrument) up to 5 years post-surgery. QoL trajectories were modelled using latent class growth analysis. Quality-adjusted life-years (QALYs) were calculated to illustrate longer term health benefit. Multinomial logistic regression analyses were performed to examine the association between trajectory groups and baseline patient characteristics.

RESULTS

After exclusions, 1553 patients out of 1892 were included in the analysis. Six unique QoL trajectories were identified; with differing levels at baseline and improvement patterns post-surgery. Only 18.4% of patients were identified to be in the most positive QoL trajectory (low baseline, large sustainable improvement after surgery) associated with the greatest gain in QALY. These patients were likely to be younger, have no co-morbidities and report greater pain at pre-surgery than most in other QoL trajectories.

CONCLUSIONS

Our findings demonstrate the importance of underlying heterogeneity in QoL trajectories, resulting in variable QALY gains. There is scope in translating routinely collected PROMs to improve shared decision-making allowing for more patient engagement. However, further research is required to identify suitable approaches of its implementation into practice to guide clinical care and maximize patient outcomes.

The WOMAC Pain Scale and Crosstalk From Co-occurring Pain Sites in People With Knee Pain: A Causal Modeling Study

OBJECTIVE

The Western Ontario and McMaster Universities Osteoarthritis (WOMAC) pain scale quantifies knee pain severity with activities of daily living, but the potential impact of pain in other body regions on WOMAC pain scores has not been explored using a causal modeling approach. The purpose of this study was to determine if pain in other areas of the body impact WOMAC pain scores, a phenomenon referred to as "crosstalk."

METHODS

Cross-sectional datasets were built from public use data available from the Osteoarthritis Initiative (OAI) and the Multicenter Osteoarthritis Study (MOST). The WOMAC Pain Scale and generic hip, knee, ankle, foot and back pain measures were included. Three nested regression models grounded in causally based classical test theory determined the extent of crosstalk. Improvements in the coefficient of determination across the 3 models were used to determine the presence of crosstalk.

RESULTS

Causal modeling provided evidence of crosstalk in both OAI and MOST datasets. For example, in OAI, multiple statistical models demonstrated significant increases in coefficient of determination values (P < .0001) as additional pain areas were added to the models.

CONCLUSIONS

Crosstalk appears to be a clinically important source of error in the WOMAC Pain Scale, particularly for patients with a larger number of painful body regions and when contralateral knee joint pain is more severe.

IMPACT STATEMENT

This study has important implications for arthritis research. It also should raise clinician awareness of the threat to score interpretation and the need to consider the extent of pain in other body regions when interpreting WOMAC pain scores.

Exploring the Impact of Quality of Life on Survival: A Case Study in Total Knee Replacement Surgery

Background. There is growing evidence that quality of life (QoL) has a strong association with mortality. However, incorporation of QoL is uncommon in standard survival modeling. Methods. Using data extracted from a registry of patients undergoing total knee replacement (TKR), the impact of incorporating QoL in survival modeling was explored using 4 parametric survival models. QoL was incorporated and tested in 2 forms, which are baseline and change in QoL due to intervention. Life expectancy and quality-adjusted life years (QALYs) were calculated and comparisons made to a reference model (no QoL) to translate the findings in the context of modeled economic evaluations. Results. A total of 2858 TKR cases (2309 patients) who had TKR between 2006 and 2015 were included in this analysis. Increases in baseline and change in QoL were associated with a reduction in mortality. Compared to the reference model, differences of up to 0.32 life years and 0.53 QALYs were observed, and these translated into a 9.5% change in incremental effectiveness. These differences were much larger as the strength of the association between QoL and mortality increased. Conclusions. This work has demonstrated that the inclusion of QoL measures (at baseline and change from baseline) when extrapolating survival does matter. It can influence health outcomes such as life expectancy and QALYs, which are relevant in cost-effectiveness analysis. This is important because neglecting the correlation between QoL and mortality can lead to imprecise extrapolations and thus risk misleading results affecting subsequent decisions made by policy makers.

Management of obstructive sleep apnoea in a primary care vs sleep unit setting: a randomised controlled trial

Objective

To assess the effectiveness and cost-effectiveness of primary care (PC) and sleep unit (SU) models for the management of subjects with suspected obstructive sleep apnoea (OSA).

Methods

Multicentre, open-label, two-arm, parallel-group, non-inferiority randomised controlled trial. A total of 302 subjects with suspected OSA and/or resistant hypertension were consecutively enrolled, 149 were treated at 11 PC units and 153 patients at a SU. The primary outcomes were a 6-month change in the Epworth Sleepiness Scale (ESS) score and Health Utilities Index (HUI). The non-inferiority margin for the ESS score was −2.0.

Results

A total of 80.2% and 70.6% of the PC and SU patients were diagnosed with OSA, respectively, and 59.3% and 60.4% of those were treated with CPAP in PC and SU units, respectively. The Apnoea–Hypopnoea Index was similar between the groups (PC vs SU (median (IQR); 23.1 (26.8) events/h vs 21.8 (35.2) events/h), and the baseline ESS score was higher in the PC than in the SU group (10.3 (6.6) vs 9 (7.2)). After 6 months, the ESS score of the PC group decreased from a mean of 10.1 to 7.6 (−2.49; 95% CI −3.3 to −1.69), and that of the SU group decreased from 8.85 to 5.73 (−3.11; 95% CI −3.94 to 2.28). The adjusted difference between groups for the mean change in the ESS score was −1.25 (one-sided 95% CI −1.88; p=0.025), supporting the non-inferiority of PC management. We did not observe differences in the HUI between groups. The cost analysis showed a median savings of €558.14/patient for the PC setting compared with the SU setting.

Conclusions

Among patients with suspected OSA, the PC model did not result in a worse ESS score or HUI than the specialist model and generated savings in terms of management cost. Therefore, the PC model was more cost-efficient than the SU model.

Trial registration

Results; >>NCT02234765, Clinical Trials.gov.

Intra-articular hyaluronic acid injection in treating knee osteoarthritis: assessing risk of bias in systematic reviews with ROBIS tool

OBJECTIVES

Intra-articular injection of hyaluronic acid (HA) is a common, yet controversial therapeutic option in treating knee osteoarthritis (OA). The purpose of the present study was to assess the risk of bias (RoB) of systematic reviews (SRs) and to summarize available evidence of HA in treating knee OA.

METHODS

A systematic search of SRs published through to December 2016 was conducted using the MEDLINE, EMBASE and Cochrane Library. The RoB of included SRs was assessed by ROBIS tool. In addition, the methodological quality of primary studies in SRs with low RoB was evaluated according to the Cochrane Handbook. The evidence quality of each primary outcome of SRs with low RoB was determined by the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) system.

RESULTS

Thirty-one SRs were eligible for inclusion. According to the ROBIS tool, there were 13 SRs with low RoB, 16 with high RoB and two with unclear RoB. The methodological quality of a total of 135 primary studies was evaluated and summarized. Forty-two outcomes from these 13 SRs were classified into the four following quality levels based on the GRADE approach: three outcomes with high quality, eight with moderate quality, 12 with low quality and 19 with very low quality.

CONCLUSIONS

This study evaluated RoB in SRs for managing knee OA with HA and assessed the evidence quality of each primary outcome in SRs with low RoB. These results can help users of SRs to improve the process of SR assessment in developing overviews or guidelines, leading to more reliable recommendations for improvements in treating knee OA. Registration: PROSPERO ((http://www.crd.york.ac.uk/PROSPERO) [CRD42017057384].