Effectiveness of Pay-for-Performance Incentive Designs on Diabetes Care

Assessing Taiwan's pay-for-performance program for diabetes care: a cost-benefit net value approach

Pay-for-Performance (P4P) to better manage chronic conditions has yielded mixed results. A better understanding of the cost and benefit of P4P is needed to improve program assessment. To this end, we assessed the effect of a P4P program using a quasi-experimental intervention and control design. Two different intervention groups were used, one consisting of newly enrolled P4P patients, and another using P4P patients who have been enrolled since the beginning of the study. Patient-level data on clinical indicators, utilization and expenditures, linked with national death registry, were collected for diabetic patients at a large regional hospital in Taiwan between 2007 and 2013. Net value, defined as the value of life years gained minus the cost of care, is calculated and compared for the intervention group of P4P patients with propensity score-matched non-P4P samples. We found that Taiwan's implementation of the P4P program for diabetic care yielded positive net values, ranging from $40,084 USD to $348,717 USD, with higher net values in the continuous enrollment model. Our results suggest that the health benefits from P4P enrollment may require a sufficient time frame to manifest, so a net value approach incorporating future predicted mortality risks may be especially important for studying chronic disease management. Future research on the mechanisms by which the Taiwan P4P program helped improve outcomes could help translate our findings to other clinical contexts.

Does the implementation of an incentive scheme increase adherence to diabetes guidelines? A retrospective cohort study of managed care enrollees

BACKGROUND

A novel incentive scheme based on a joint agreement of a large Swiss health insurance with 56 physician networks was implemented in 2018. This study evaluated the effect of its implementation on adherence to evidence-based guidelines among patients with diabetes in managed care models.

METHODS

We performed a retrospective cohort study, using health care claims data from patients with diabetes enrolled in a managed care plan (2016-2019). Guideline adherence was assessed by four evidence-based performance measures and four hierarchically constructed adherence levels. Generalized multilevel models were used to examine the effect of the incentive scheme on guideline adherence.

RESULTS

A total of 6'273 patients with diabetes were included in this study. The raw descriptive statistics showed minor improvements in guideline adherence after the implementation. After adjusting for underlying patient characteristics and potential differences between physician networks, the likelihood of receiving a test was moderately but consistently higher after the implementation of the incentive scheme for most performance measures, ranging from 18% (albuminuria: OR, 1.18; 95%-CI, 1.05-1.33) to 58% (HDL cholesterol: OR, 1.58; 95%-CI, 1.40-1.78). Full adherence was more likely after implementation of the incentive scheme (OR, 1.37; 95%-CI, 1.20-1.55), whereas level 1 significantly decreased (OR, 0.74; 95%-CI, 0.65 - 0.85). The proportions of the other adherence levels were stable.

CONCLUSION

Incentive schemes including transparency of the achieved performance may be able to improve guideline adherence in patients with diabetes and are promising to increase quality of care in this patient population.

Effect of a Pay-for-Performance Program on Renal Outcomes Among Patients With Early-Stage Chronic Kidney Disease in Taiwan

BACKGROUND

With the promising outcomes of the pre-ESRD (end-stage renal disease) pay-for-performance (P4P) program, the National Health Insurance Administration (NHIA) of Taiwan launched a P4P program for patients with early chronic kidney disease (CKD) in 2011, targeting CKD patients at stages 1, 2, and 3a. This study aimed to examine the long-term effect of the early-CKD P4P program on CKD progression.

METHODS

We conducted a matched cohort study using electronic medical records from a large healthcare delivery system in Taiwan. The outcome of interest was CKD progression to estimated glomerular filtration rate (eGFR) <45 mL/min/1.73 m2 between P4P program enrolees and non-enrolees. The difference in the cumulative incidence of CKD progression between the P4P and non-P4P groups was tested using Gray's test. We adopted a cause-specific (CS) hazard model to estimate the hazard in the P4P group as compared to non-P4P group, adjusting for age, sex, baseline renal function, and comorbidities. A subgroup analysis was further performed in CKD patients with diabetes to evaluate the interactive effects between the early-CKD P4P and diabetes P4P programs.

RESULTS

The incidence per 100 person-months of disease progression was significantly lower in the P4P group than in the non-P4P group (0.44 vs. 0.69, P<.0001), and the CS hazard ratio (CS-HR) for P4P program enrolees compared with non-enrolees was 0.61 (95% CI: 0.58-0.64, P<.0001). The results of the subgroup analysis further revealed an additive effect of the diabetes P4P program on CKD progression; compared to none of both P4P enrolees, the CS-HR for CKD disease progression was 0.60 (95% CI: 0.54-0.67, P<.0001) for patients who were enrolled in both early-CKD P4P and diabetes P4P programs.

CONCLUSION

The present study results suggest that the early-CKD P4P program is superior to usual care to decelerate CKD progression in patients with early-stage CKD.

Relationships Between Bronchodilators, Steroids, Antiarrhythmic Drugs, Antidepressants, and Benzodiazepines and Heart Disease and Ischemic Stroke in Patients With Predominant Bronchiectasis and Asthma

Objective

We investigated the effects of medication on heart disease and ischemic stroke (HDS) risk in patients with predominant bronchiectasis-asthma combination (BCAS).

Methods

BCAS and non-BCAS cohorts (N = 588 and 1,118, respectively) were retrospectively enrolled. The cumulative incidence of HDS was analyzed using Cox proportional regression; propensity scores were estimated using non-parsimonious multivariable logistic regression. Adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) for HDS were calculated, adjusting for sex, age, comorbidities, and medication {long- and short-acting β2 agonists and muscarinic antagonists (LABAs/SABAs and LAMAs/SAMAs), steroids [inhaled corticosteroid steroids (ICSs), oral steroids (OSs)], antiarrhythmics, antidepressants (fluoxetine), benzodiazepines (alprazolam, fludiazepam), statins and antihypertensive drugs (diuretics, cardioselective beta blockers, calcium channel blockers (CCBs) and angiotensin converting enzyme inhibitors (ACEi), angiotensin II blockers)}.

Results

Compared with the non-BCAS cohort, the BCAS cohort taking LABAs, SABAs, SAMAs, ICSs, OSs, antiarrhythmics, and alprazolam had an elevated HDS risk [aHRs (95% CIs): 2.36 (1.25–4.33), 2.65 (1.87–3.75), 2.66 (1.74–4.05), 2.53 (1.61–3.99), 1.76 (1.43–2.18), 9.88 (3.27–30.5), and 1.73 (1.15–2.58), respectively except fludiazepam 1.33 (0.73–2.40)]. The aHRs (95% CIs) for LABAs ≤ 30 days, DDDs &lt;415, ICSs ≤ 30 days were 1.10 (0.38–3.15), 2.95 (0.22–38.8), 1.45 (0.76–2.77). The aHRs (95% CIs) for current and recent alprazolam were 1.78 (1.09–2.93) and 777.8 (1.34–451590.0); for current and past fludiazepam were 1.39 (0.75–2.59) and 1.29 (0.42–4.01) and for past alprazolam was 1.57 (0.55–4.46); respectively. The aHRs (95% CIs) for alprazolam &gt;30 DDDs, fludiazepam &gt;20 DDDs, ICSs ≦415 DDDs, and OSs DDDs ≦15 were 1.60 (0.78–3.29), 2.43 (0.90–6.55), 5.02 (1.76–14.3), and 2.28 (1.43–3.62), respectively.

Conclusion

The bronchodilators, steroids, and antiarrhythmics were associated with higher risk of HDS, even low dose use of steroids. However, the current use of LABAs/ICSs were not associated with HDS. Benzodiazepines were relatively safe, except for current or recent alprazolam use. Notably, taking confounders into account is crucial in observational studies.

The Effectiveness of Pay-for-Performance Program of Diabetes Care for Psychiatric Patients: A Regional Psychiatric Hospital Experience

BACKGROUND

The Pay-for-performance (P4P) program of diabetes care has demonstrated successful outcomes in patients with type 2 diabetes. However, the effectiveness of this multidisciplinary care model for psychiatric patients has never been evaluated. The objective of this study is to examine the effectiveness of P4P program of diabetes for psychiatric patients with diabetes.

METHODS

This study utilized a retrospective cohort design to examine the effectiveness of P4P program of diabetes care for psychiatric patients with diabetes. The participants' HbA1c (hemoglobin A1c) data of the fourth quarter in 2018 were used as baseline value, while P4P program was not applied yet. HbA1c data of every quarter in 2019 were collected. Generalized estimating equations (GEE) was used to analyze the change of HbA1c level.

RESULTS

The HbA1c level increased slightly in the first quarter, and then decreased gradually since the second quarter. The HbA1c level was significantly lower in the fourth quarter after P4P program intervention (p < 0.05).

CONCLUSION

P4P program of diabetes care is also effective on psychiatric patients with diabetes, and this multidisciplinary care model could be encouraged and promoted for psychiatric patients with diabetes.

The Linkages Between Reimbursement and Prevention: A Mixed-Methods Approach

Background: The benefits of prevention are widely recognized; ranging from avoiding disease onset to substantially reducing disease burden, which is especially relevant considering the increasing prevalence of chronic diseases. However, its delivery has encountered numerous obstacles in healthcare. While healthcare professionals play an important role in stimulating prevention, their behaviors can be influenced by incentives related to reimbursement schemes.

Purpose: The purpose of this research is to obtain a detailed description and explanation of how reimbursement schemes specifically impact primary, secondary, tertiary, and quaternary prevention.

Methods: Our study takes a mixed-methods approach. Based on a rapid review of the literature, we include and assess 27 studies. Moreover, we conducted semi-structured interviews with eight Dutch healthcare professionals and two representatives of insurance companies, to obtain a deeper understanding of healthcare professionals' behaviors in response to incentives.

Results: Nor fee-for-service (FFS) nor salary can be unambiguously linked to higher or lower provision of preventive services. However, results suggest that FFS's widely reported incentive to increase production might work in favor of preventive services such as immunizations but provide less incentives for chronic disease management. Salary's incentive toward prevention will be (partially) determined by provider-organization's characteristics and reimbursement. Pay-for-performance (P4P) is not always necessarily translated into better health outcomes, effective prevention, or adequate chronic disease management. P4P is considered disruptive by professionals and our results expose how it can lead professionals to resort to (over)medicalization in order to achieve targets. Relatively new forms of reimbursement such as population-based payment may incentivize professionals to adapt the delivery of care to facilitate the delivery of some forms of prevention.

Conclusion: There is not one reimbursement scheme that will stimulate all levels of prevention. Certain types of reimbursement work well for certain types of preventive care services. A volume incentive could be beneficial for prevention activities that are easy to specify. Population-based capitation can help promote preventive activities that require efforts that are not incentivized under other reimbursements, for instance activities that are not easily specified, such as providing education on lifestyle factors related to a patient's (chronic) disease.

Is the evidence on the effectiveness of pay for performance schemes in healthcare changing? Evidence from a meta-regression analysis

BACKGROUND

This study investigated if the evidence on the success of the Pay for Performance (P4P) schemes in healthcare is changing as the schemes continue to evolve by updating a previous systematic review.

METHODS

A meta-regression analysis using 116 studies evaluating P4P schemes published between January 2010 to February 2018. The effects of the research design, incentive schemes, use of incentives, and the size of the payment to revenue ratio on the proportion of statically significant effects in each study were examined.

RESULTS

There was evidence of an increase in the range of countries adopting P4P schemes and weak evidence that the proportion of studies with statistically significant effects have increased. Factors hypothesized to influence the success of schemes have not changed. Studies evaluating P4P schemes which made payments for improvement over time, were associated with a lower proportion of statistically significant effects. There was weak evidence of a positive association between the incentives' size and the proportion of statistically significant effects.

CONCLUSION

The evidence on the effectiveness of P4P schemes is evolving slowly, with little evidence that lessons are being learned concerning the design and evaluation of P4P schemes.

A retrospective cohort study on the risk of stroke in relation to a priori health knowledge level among people with type 2 diabetes mellitus in Taiwan

Background

Intervention of diabetes care education with regular laboratory check-up in outpatient visits showed long-term benefits to reduce the risk of macrovascular complications among people with type 2 diabetes. However, research on the level of a priori health knowledge to the prevention of diabetic complications in community settings has been scarce. We therefore aimed to investigate the association of health knowledge and stroke incidence in patients with type 2 diabetes in Taiwan.

Methods

A nationally representative sample of general Taiwanese population was selected using a multistage systematic sampling process from Taiwan National Health Interview Survey (NHIS) in 2005. Subjects were interviewed by a standardized face-to-face questionnaire in the survey, obtaining information of demographics, socioeconomic status, family medical history, obesity, health behaviors, and 15-item health knowledge assessment. The NHIS dataset was linked to Taiwan National Health Insurance claims data to retrieve the diagnosis of type 2 diabetes in NHIS participants at baseline and identify follow-up incidence of stroke from 2005 to 2013. Univariate and multivariate Cox regressions were used to estimate the effect of baseline health knowledge level to the risk of stroke incidence among this group of people with type 2 diabetes.

Results

A total of 597 diabetic patients with a mean age of 51.28 years old and nearly half of males were analyzed. During the 9-year follow-up period, 65 new stroke cases were identified among them. Kaplan–Meier curves comparing the three groups of low/moderate/high knowledge levels revealed a statistical significance (p-value of log-rank test <0.01). After controlling for potential confounders, comparing to the group of low health knowledge level, the relative risk of stroke was significantly lower for those with moderate (adjusted hazard ratio [AHR] = 0.63; 95% CI, 0.33–1.19; p-value = 0.15) and high level of health knowledge (AHR = 0.43; 95% CI, 0.22–0.86; p-value = 0.02), with a significant linear trend (p-value = 0.02).

Conclusions

An exposure-response gradient of moderate to high health knowledge levels to the prevention of stroke incidence among people with type 2 diabetes in community was found with 9 years of follow-up in Taiwan. Development and delivery of health education on stroke prevention to people with type 2 diabetes are warranted.