Background

The Patient-Centered Medical Home (PCMH) is a widely endorsed model of delivery system reform that emphasizes primary care. Pilot demonstration projects are underway in many states, sponsored by Medicare, Medicaid, major health plans and multi-payer coalitions.

Methods

In this paper we consider the development of a long-term policy-relevant research agenda on outcomes of the PCMH. We provide an overview of potential measures of PCMH impact, identify measurement challenges and recommend areas for further study. Although the PCMH should not be expected to solve every problem in the health care system, developing a research agenda for measuring outcomes of delivery system innovations such as the PCMH should be considered in the context of the larger effort to improve the US health care system, with the ultimate goal to improve population health.

Results

As a framework for our discussion, we have chosen the Institute of Medicine's six specific aims for 21st century health care: (1) safe, (2) effective, (3) patient-centered, (4) timely, (5) efficient and (6) equitable. In addition, we include potential areas of PCMH outcomes that do not easily fall under this framework and consider unintended consequences.

Conclusion

Multi-stakeholder involvement will be essential in developing a long-term policy-relevant research agenda for outcomes of the PCMH.

Context

Research on predictors of clinical outcomes usually focuses on the impact of individual patient factors, despite known relationships between neighborhood environment and health.

Objective

To determine whether US census information on where a patient resides is associated with diabetes development among patients with prediabetes.

Design

Retrospective cohort study of all 157,752 patients aged 18 years or older from Kaiser Permanente Northern California with laboratory-defined prediabetes (fasting plasma glucose, 100 mg/dL-125 mg/dL, and/or glycated hemoglobin, 5.7%-6.4%). We assessed whether census data on education, income, and percentage of households receiving benefits through the US Department of Agriculture's Supplemental Nutrition Assistance Program (SNAP) was associated with diabetes development using logistic regression controlling for age, sex, race/ethnicity, blood glucose levels, and body mass index.

Main outcome measure

Progression to diabetes within 36 months.

Results

Patients were more likely to progress to diabetes if they lived in an area where less than 16% of adults had obtained a bachelor's degree or higher (odds ratio [OR] =1.22, 95% confidence interval [CI] = 1.09-1.36), where median annual income was below $79,999 (OR = 1.16 95% CI = 1.03-1.31), or where SNAP benefits were received by 10% or more of households (OR = 1.24, 95% CI = 1.1-1.4).

Conclusion

Area-level socioeconomic and food assistance data predict the development of diabetes, even after adjusting for traditional individual demographic and clinical factors. Clinical interventions should take these factors into account, and health care systems should consider addressing social needs and community resources as a path to improving health outcomes.

The Patient-Centered Medical Home (PCMH) is a widely endorsed model of delivery system reform that emphasizes primary care. Pilot demonstration projects are underway in many states, sponsored by Medicare, Medicaid, major health plans and multi-payer coalitions. In this paper we consider the development of a long-term policy-relevant research agenda on outcomes of the PCMH. We provide an overview of potential measures of PCMH impact, identify measurement challenges and recommend areas for further study. Although the PCMH should not be expected to solve every problem in the health care system, developing a research agenda for measuring outcomes of delivery system innovations such as the PCMH should be considered in the context of the larger effort to improve the US health care system, with the ultimate goal to improve population health. As a framework for our discussion, we have chosen the Institute of Medicine’s six specific aims for 21st century health care: (1) safe, (2) effective, (3) patient-centered, (4) timely, (5) efficient and (6) equitable. In addition, we include potential areas of PCMH outcomes that do not easily fall under this framework and consider unintended consequences. Multi-stakeholder involvement will be essential in developing a long-term policy-relevant research agenda for outcomes of the PCMH.

Background

Shared decision making (SDM) is a potentially valuable but underexplored approach to contraceptive counseling.

Methods

We determined the correlation between patient report of SDM and of whether their provider had a method preference with measures of satisfaction.

Results

Women reporting SDM were more likely to be satisfied with counseling than those reporting a provider-driven decision and were more likely to be satisfied with their method than those reporting a patient-driven decision. Patients who felt that the provider had a method preference were less likely to be satisfied with their method.

Conclusions

SDM in contraceptive counseling is associated with patient satisfaction.

Objectives

Provider self-disclosure (PSD) - defined as providers making statements regarding personal information to patients - has not been well characterized in the context of contraceptive counseling. In this study, we describe the incidence, content and context of contraceptive PSD.

Study design

This mixed methods analysis used data from the Provider-Patient Contraceptive Counseling study, for which 349 family planning patients were recruited from 2009 to 2012 from six clinics in the San Francisco Bay Area. Audio-recordings from their visits were analyzed for the presence or absence of PSD, and those visits with evidence of PSD were analyzed using qualitative methods. The associations of patient and provider demographics and patient satisfaction measures, obtained from survey data, with PSD were analyzed using bivariable and multivariable analyses.

Results

Thirty-seven percent of providers showed evidence of PSD during at least one visit, and PSD occurred in 9% of clinic visits. Fifty-four percent of PSD statements were about intrauterine devices. About half of PSD statements occurred prior to the final selection of the contraceptive method and appeared to influence the choice of method. In post-visit surveys, all patients who reported receiving PSD considered it to be appropriate, and patient-reported PSD was not statistically associated with measures of patient satisfaction.

Conclusions

This study provides some support for the appropriateness of PSD during family planning encounters, at least as practiced during the sampled visits. Further research could explore whether this counseling strategy has an impact on patients' ability to identify the best contraceptive methods for them.

Implications

In this study, PSD did not have a demonstrated negative effect on the provider-patient relationship. In almost half of visits, PSD appeared to influence patients' choice of a method; whether this influence is beneficial needs further research.

Objective

Consistent estimation of causal effects with inverse probability weighting estimators is known to rely on consistent estimation of propensity scores. To alleviate the bias expected from incorrect model specification for these nuisance parameters in observational studies, data-adaptive estimation and in particular an ensemble learning approach known as Super Learning has been proposed as an alternative to the common practice of estimation based on arbitrary model specification. While the theoretical arguments against the use of the latter haphazard estimation strategy are evident, the extent to which data-adaptive estimation can improve inferences in practice is not. Some practitioners may view bias concerns over arbitrary parametric assumptions as academic considerations that are inconsequential in practice. They may also be wary of data-adaptive estimation of the propensity scores for fear of greatly increasing estimation variability due to extreme weight values. With this report, we aim to contribute to the understanding of the potential practical consequences of the choice of estimation strategy for the propensity scores in real-world comparative effectiveness research.

Method

We implement secondary analyses of Electronic Health Record data from a large cohort of type 2 diabetes patients to evaluate the effects of four adaptive treatment intensification strategies for glucose control (dynamic treatment regimens) on subsequent development or progression of urinary albumin excretion. Three Inverse Probability Weighting estimators are implemented using both model-based and data-adaptive estimation strategies for the propensity scores. Their practical performances for proper confounding and selection bias adjustment are compared and evaluated against results from previous randomized experiments.

Conclusion

Results suggest both potential reduction in bias and increase in efficiency at the cost of an increase in computing time when using Super Learning to implement Inverse Probability Weighting estimators to draw causal inferences.

Background

Contraceptive nonadherence is an important contributor to unintended pregnancy in the United States. While the elimination of patient cost sharing has been cited as means to improve contraceptive access, little is known about the relationship between cost sharing and ongoing adherence and continuation of chosen methods. The purpose of this study was to examine the relationship between copayment amount and adherence to pharmacy-dispensed contraception in young women.

Methods

We conducted a retrospective cohort study of 39,142 women ages 19-29 with a new prescription for the contraceptive pill, patch, or ring at Kaiser Permanente Northern California during 2011-2014. We examined 12-month nonadherence as measured by timely prescription refills and used multivariable Cox proportional hazards models to assess the association between copayment amount and the risk of nonadherence.

Results

Ninety-four percent of women used the pill, and 6% used the patch or ring. Forty percent of patients had no copayment and 25% had a copayment of ≥$30. Nearly 75% of women were nonadherent during the study period. In 2013 and 2014, women with a copayment had a 9% increased risk of nonadherence (adjusted hazard ratio, 1.09; 95% confidence interval, 1.04, 1.14) compared with women with no copayment.

Conclusions

Prescription copayments may serve as a barrier to adherence of pharmacy-dispensed contraception. Given recent changes to Affordable Care Act contraceptive coverage requirement, these findings can be used to support state-level and health system-level policies for no-cost contraception, and to determine the potential public health impact of this policy change.

Objective

To identify positive deviant (PD) physician organizations of Accountable Care Organizations (ACOs) with robust performance management systems (PMSYS).

Data source

Third National Survey of Physician Organizations (NSPO3, n = 1,398).

Study design

Organizational and external factors from NSPO3 were analyzed.

Data collection/extraction methods

Linear regression estimated the association of internal and contextual factors on PMSYS. Two cutpoints (75th/90th percentiles) identified PDs with the largest residuals and highest PMSYS scores.

Principal findings

A total of 65 and 41 PDs were identified using 75th and 90th percentiles cutpoints, respectively. The 90th percentile more strongly differentiated PDs from non-PDs. Having a high proportion of vulnerable patients appears to constrain PMSYS development.

Conclusions

Our PD identification method increases the likelihood that PD organizations selected for in-depth inquiry are high-performing organizations that exceed expectations.

Patient–physician race/ethnicity and language concordance may improve medication adherence and reduce disparities in cardiovascular disease (CVD) by fostering trust and improved patient–physician communication. To examine the association of patient race/ethnicity and language and patient–physician race/ethnicity and language concordance on medication adherence rates for a large cohort of diabetes patients in an integrated delivery system. We studied 131,277 adult diabetes patients in Kaiser Permanente Northern California in 2005. Probit models assessed the effect of patient and physician race/ethnicity and language on adherence to CVD medications, after controlling for patient and physician characteristics. Ten percent of African American, 11 % of Hispanic, 63% of Asian, and 47% of white patients had same race/ethnicity physicians. 24% of Spanish-speaking patients were linguistically concordant with their physicians. African American (46%), Hispanic (49%) and Asian (52%) patients were significantly less likely than white patients (58%) to be in good adherence to all of their CVD medications (p < 0.001). Spanish-speaking patients were less likely than English speaking patients to be in good adherence (51% versus 57%, p < 0.001). Race concordance for African American patients was associated with adherence to all their CVD medications (53% vs. 50%, p < 0.05). Language concordance was associated with medication adherence for Spanish-speaking patients (51% vs. 45%, p < 0.05). Increasing opportunities for patient–physician race/ethnicity and language concordance may improve medication adherence for African American and Spanish-speaking patients, though a similar effect was not observed for Asian patients or English-proficient Hispanic patients.

Objective

To examine the predictors of patient-physician race/ethnicity concordance among diabetes patients in an integrated delivery system.

Data source

Kaiser Permanente's Northern California Diabetes Registry of 2005.

Study design

Logistic regression predicted concordance for each racial/ethnic group. Availability of a concordant physician, whether a patient chose their physician, and patient language were main explanatory variables.

Data collection/extraction methods

The study population consisted of 109,745 patients and 1,750 physicians.

Principal findings

Patients who chose their physicians were more likely to have a same race/ethnicity physician with OR of 2.2 (95 percent CI 1.74-2.82) for African American patients, 1.71 (95 percent CI 1.44-2.04) for Hispanic patients, 1.11 (95 percent CI 1.04-1.18) for white patients, and 1.38 (95 percent CI 1.23, 1.55) for Asian patients. Availability of a same race/ethnicity physician was also a predictor of concordance for African American patients (OR 2.7; 95 percent CI 2.45-2.98) and marginally significant for Hispanic patients (OR 1.02; 95 percent CI 1.01-1.02), white patients (OR 1.02; 95 percent CI 1.00-1.04), and Asian patients (OR 1.05; 95 percent CI 1.03, 1.07). Limited English language was a strong predictor of concordance for Hispanic patients (OR 4.81; 95 percent CI 4.2-5.51) and Asian patients (OR 9.8; 95 percent CI 7.7, 12.6).

Conclusion

Patient language, preferences, and the racial composition of the physician workforce predict race/ethnicity concordance.