Identification and classification of principal features for analyzing unwarranted clinical variation

RATIONALE, AIMS, AND OBJECTIVE

Unwarranted clinical variation (UCV) is an undesirable aspect of a healthcare system, but analyzing for UCV can be difficult and time-consuming. No analytic feature guidelines currently exist to aid researchers. We performed a systematic review of UCV literature to identify and classify the features researchers have identified as necessary for the analysis of UCV.

METHODS

The literature search followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses. We looked for articles with the terms 'medical practice variation' and 'unwarranted clinical variation' from four databases: Medline, Web of Science, EMBASE and CINAHL. The search was performed on 24 March 2023. The articles selected were original research articles in the English language reporting on UCV analysis in adult populations. Most of the studies were retrospective cohort analyses. We excluded studies reporting geographic variation based on the Atlas of Variation or small-area analysis methods. We used ASReview Lab software to assist in identifying articles for abstract review. We also conducted subsequent reference searches of the primary articles to retrieve additional articles.

RESULTS

The search yielded 499 articles, and we reviewed 46. We identified 28 principal analytic features utilized to analyze for unwarranted variation, categorised under patient-related or local healthcare context factors. Within the patient-related factors, we identified three subcategories: patient sociodemographics, clinical characteristics, and preferences, and classified 17 features into seven subcategories. In the local context category, 11 features are classified under two subcategories. Examples are provided on the usage of each feature for analysis.

CONCLUSION

Twenty-eight analytic features have been identified, and a categorisation has been established showing the relationships between features. Identifying and classifying features provides guidelines for known confounders during analysis and reduces the steps required when performing UCV analysis; there is no longer a need for a UCV researcher to engage in time-consuming feature engineering activities.

Medical Practice Variation Among Primary Care Physicians: 1 Decade, 14 Health Services, and 3,238,498 Patient-Years

PURPOSE

Variation in medical practice is associated with poorer health outcomes, increased costs, disparities in care, and increased burden on the public health system. In the present study, we sought to describe and assess inter- and intra-primary care physician variation, adjusted for patient and clinic characteristics, over a decade of practice and across a broad range of health services.

METHODS

We assessed practice patterns of 251 primary care physicians in southern Israel. For each of 14 health services (imaging tests, cardiac tests, laboratory tests, and specialist visits) we described interphysician and intraphysician variation, adjusted for patient case mix and clinic characteristics, using the coefficient of variation. The adjusted rates were assessed by generalized linear negative-binomial mixed models.

RESULTS

The variation between physicians was on average 3-fold greater than the variation of individual physician practice over the years. Services with low utilization were associated with greater inter- and intraphysician variation: r_s = (-0.58), P = .03 and r_s = (-0.39), P = .17, respectively. In addition, physician utilization ranks averaged over all health services were consistent across the 14 health services (intraclass correlation coefficient, 0.94; 95% CI, 0.93-0.95).

CONCLUSIONS

Our results show greater variation in practice patterns between physicians than for individual physicians over the years. It appears that the variation remains high even after adjustment for patient and clinic characteristics and that the individual physician utilization patterns are stable across health services. We propose that personal behavioral characteristics of medical practitioners might explain this variation.

Determinants of Medical Practice Variation Among Primary Care Physicians: Protocol for a Three Phase Study

BACKGROUND

One of the greatest challenges of modern health systems is the choice and use of resources needed to diagnose and treat patients. Medical practice variation (MPV) is a broad term which entails the differences between health care providers inclusive of both the overuse and underuse. In this paper, we describe a 3-phase research protocol examining MPV in primary care.

OBJECTIVE

We aim to identify the potential targets for behavioral modification interventions to reduce the variation in practice patterns and thus improve health care, decrease costs, and prevent disparities in care.

METHODS

The first phase will delineate the variation in primary care practice over a wide range of services and long follow-up period (2003-2017), the second will examine the 3 determinants of variation (ie, patient, physician, and clinic characteristics), and attempt to derive the unexplained variance. In the third phase, we will assess a novel component that might contribute to the previously unexplained variance - the physicians' personal behavioral characteristics (such as risk aversion, fear of malpractice, stress from uncertainty, empathy, and burnout).

RESULTS

This work was supported by the research grant from Israel National Institute for Health Policy Research (Grant No. 2014/134). Soroka University Medical Center Institutional Ethics Committee has approved the updated version of the study protocol (SOR-14-0063) in February 2019. All relevant data for phases 1 and 2, including patient, physician, and clinic, were collected from the Clalit Health Services data set in 2019 and are currently being analyzed. The evaluation of the individual physician characteristics (eg, risk aversion) by the face-to-face questionnaires was started on 2018 and remains in progress. We intend to publish the results during 2020-2021.

CONCLUSIONS

Based on the results of our study, we aim to propose a list of potential targets for focused behavioral intervention. Identifying new targets for such an intervention can potentially lead to a decrease in the unwarranted variation in the medical practice. We suggest that such an intervention will result in optimization of the health system, improvement of health outcomes, reduction of disparities in care and savings in cost.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/18673.

Caesarean section in uninsured women in the USA: systematic review and meta-analysis

OBJECTIVE

The aim of this study is to assess the odds of caesarean section (CS) for uninsured women in the USA and understand the underlying mechanisms as well as consequences of lower use.

STUDY DESIGN

Systematic review and meta-analysis.

DATA SOURCES

PubMed, Embase, the Cochrane Library and CINAHL from the first year of records to April 2018.

ELIGIBILITY CRITERIA

We included studies that reported data to allow the calculation of ORs of CS of uninsured as compared with insured women.

OUTCOMES

The prespecified primary outcome was the adjusted OR of deliveries by CS of uninsured women as compared with privately or publicly insured women. The prespecified secondary outcome was the crude OR of deliveries by CS of uninsured women as compared with insured women.

RESULTS

12 articles describing 16 separate studies involving more than 8.8 million women were included in this study. We found: 0.70 times lower odds of CS in uninsured as compared with privately insured women (95% CI 0.63 to 0.78), with no relevant heterogeneity between studies (τ2=0.01); and 0.92 times lower odds for CS in uninsured as compared with publicly insured women (95% CI 0.80 to 1.07), with no relevant heterogeneity between studies (τ2=0.02). We found 0.70 times lower odds in uninsured as compared with privately and publicly insured women (95% CI 0.69 to 0.72).

CONCLUSIONS

CSs are less likely to be performed in uninsured women as compared with insured women. While the higher rates for CS among privately insured women can be explained with financial incentives associated with private insurance, the lower odds among uninsured women draw attention at barriers to access for delivery care. In many regions, the rates for uninsured women are above, close or below the benchmarks for appropriate CS rates and could imply both, underuse and overuse.

Prescriber Variation in Relation to Prescribing Trends within the Preferred Drugs Initiative in Ireland (2012-2015): An Interrupted Time-Series Study Using Latent Curve Models

OBJECTIVES

To examine the impact of the Preferred Drugs Initiative (PDI), an Irish health policy aimed at reducing prescribing variation.

DESIGN

Interrupted time series spanning 2012 to 2015.

SETTING

Health Service Executive pharmacy claims data for General Medical Services (GMS) patients, approximately 40% of the Irish population.

PARTICIPANTS

Prescribers issuing preferred drug group items to GMS adults before and after PDI guidelines.

PRIMARY OUTCOME

The percentage coverage of PDI medications within each drug class per calendar quarter per prescriber.

METHODS

Latent curve models with structured residuals (LCM-SRs) were used to model coverage of the preferred drugs over time. The number of GMS adults receiving medication and the percentage who were 65 years and older at the start of the study were included as covariates.

RESULTS

In the quarter following PDI guidelines, coverage of the preferred drugs increased most in absolute terms for proton pump inhibitors (PPIs) (1.50% [SE 0.15], P < 0.001) and selective and norepinephrine reuptake inhibitors (SNRIs) (1.17% [SE 0.26], P < 0.001). Variation between prescribers remained relatively unchanged and increased for urology medications. Prescribers who increased coverage of the preferred PPI also increased coverage of the preferred statin immediately following guidelines (correlation 0.47 [SE 0.13], P < 0.001). Where guidelines were disseminated simultaneously, coverage of one preferred drug did not significantly predict coverage of the other preferred drug in the next calendar quarter. Prescribing of preferred drugs was not moderated by prescriber-level factors.

CONCLUSIONS

Modest changes in prescribing of the preferred drugs have been observed over the course of the PDI. However, the guidelines have had little impact in reducing variation between prescribers. Further strategies may be necessary to reduce variation in clinical practice and enhance patient care.

Is there a conflict between general practitioners applying guidelines for antibiotic prescribing and including their patients' preferences?

OBJECTIVES

One perceived barrier to guideline adherence is the existence of conflicting patient preferences. We examined whether patient preferences influence the prescription of antibiotics in general practice, and how this affects guideline adherence. We hypothesized that preferences play a larger role in prescribing antibiotics if the guideline allows for preferences to be taken into account, ie, if prescribing antibiotics is an option which can be considered rather than a clear recommendation to prescribe or not. We included three guidelines: acute cough, acute rhinosinusitis, and urinary tract infections.

METHODS

Data from NIVEL (the Netherlands Institute for Health Services Research) Primary Care Database (NIVEL-PCD) were used to assess antibiotic indications and prescriptions. These data were combined with a questionnaire among members of NIVEL's Dutch Health Care Consumer Panel to examine patient preferences. According to NIVEL-PCD, 286 of these members contacted their general practitioner (GP) in 2015 for acute cough, acute rhinosinusitis or urinary tract infections. A logistic multilevel regression analysis was performed to test our hypothesis.

RESULTS

Patient preferences do play a role in GPs' prescribing of antibiotics only in situations where, in accordance with the guideline, their use is an option which could be considered (interaction between indication and preference: p=0.049). If patients ask for antibiotics themselves in such situations, then GPs prescribe antibiotics more often.

CONCLUSION

Patient preferences only play a role if the guideline provides room to take preferences into account. Therefore, our results do not suggest a conflict between applying guidelines and including patient preferences. Further research is recommended to examine this possible conflict in other situations.

Caesarean sections and private insurance: systematic review and meta-analysis

OBJECTIVE

Financial incentives associated with private insurance may encourage healthcare providers to perform more caesarean sections. We therefore sought to determine the association of private insurance and odds of caesarean section.

DESIGN

Systematic review and meta-analysis.

DATA SOURCES

MEDLINE, Embase and The Cochrane Library from the first year of records through August 2016.

ELIGIBILITY CRITERIA

We included studies that reported data to allow the calculation of OR of caesarean section of privately insured as compared with publicly insured women.

OUTCOMES

The prespecified primary outcome was the adjusted OR of births delivered by caesarean section of women covered with private insurance as compared with women covered with public insurance. The prespecified secondary outcome was the crude OR of births delivered by caesarean section of women covered with private insurance as compared with women covered with public insurance.

RESULTS

Eighteen articles describing 21 separate studies in 12.9 million women were included in this study. In a meta-analysis of 13 studies, the adjusted odds of delivery by caesarean section was 1.13 higher among privately insured women as compared with women with public insurance coverage (95% CI 1.07 to 1.18) with no relevant heterogeneity between studies (τ²=0.006). The meta-analysis of crude estimates from 12 studies revealed a somewhat more pronounced association (pooled OR 1.35, 95% CI 1.27 to 1.44) with no relevant heterogeneity between studies (τ²=0.011).

CONCLUSIONS

Caesarean sections are more likely to be performed in privately insured women as compared with women using public health insurance coverage. Although this effect is small on average and variable in its magnitude, it is present in all analyses we performed.

Caesarean sections and for-profit status of hospitals: systematic review and meta-analysis

OBJECTIVE

Financial incentives may encourage private for-profit providers to perform more caesarean section (CS) than non-profit hospitals. We therefore sought to determine the association of for-profit status of hospital and odds of CS.

DESIGN

Systematic review and meta-analysis.

DATA SOURCES

MEDLINE, EMBASE and the Cochrane Database of Systematic Reviews from the first year of records through February 2016.

ELIGIBILITY CRITERIA

To be eligible, studies had to report data to allow the calculation of ORs of CS comparing private for-profit hospitals with public or private non-profit hospitals in a specific geographic area.

OUTCOMES

The prespecified primary outcome was the adjusted OR of births delivered by CS in private for-profit hospitals as compared with public or private non-profit hospitals; the prespecified secondary outcome was the crude OR of CS in private for-profit hospitals as compared with public or private non-profit hospitals.

RESULTS

15 articles describing 17 separate studies in 4.1 million women were included. In a meta-analysis of 11 studies, the adjusted odds of delivery by CS was 1.41 higher in for-profit hospitals as compared with non-profit hospitals (95% CI 1.24 to 1.60) with no relevant heterogeneity between studies (τ²≤0.037). Findings were robust across subgroups of studies in stratified analyses. The meta-analysis of crude estimates from 16 studies revealed a somewhat more pronounced association (pooled OR 1.84, 95% CI 1.49 to 2.27) with moderate-to-high heterogeneity between studies (τ²≥0.179).

CONCLUSIONS

CS are more likely to be performed by for-profit hospitals as compared with non-profit hospitals. This holds true regardless of women's risk and contextual factors such as country, year or study design. Since financial incentives are likely to play an important role, we recommend examination of incentive structures of for-profit hospitals to identify strategies that encourage appropriate provision of CS.

Predictors of adjuvant treatment for pancreatic adenocarcinoma at the population level

BACKGROUND

In the present study, we aimed to describe, at the population level, patterns of adjuvant treatment use after curative-intent resection for pancreatic adenocarcinoma (pcc) and to identify independent predictors of adjuvant treatment use.

METHODS

In this observational cohort study, patients undergoing pcc resection in the province of Ontario (population 13 million) during 2005-2010 were identified using the provincial cancer registry and were linked to administrative databases that include all treatments received and outcomes experienced in the province. Patients were defined as having received chemotherapy (ctx), chemoradiation (crt), or observation (obs). Clinicopathologic factors associated with the use of ctx, crt, or obs were identified by chi-square test. Logistic regression analyses were used to identify independent predictors of adjuvant treatment versus obs, and ctx versus crt.

RESULTS

Of the 397 patients included, 75.3% received adjuvant treatment (27.2% crt, 48.1% ctx) and 24.7% received obs. Within a single-payer health care system with universal coverage of costs for ctx and crt, substantial variation by geographic region was observed. Although the likelihood of receiving adjuvant treatment increased from 2005 to 2010 (p = 0.002), multivariate analysis revealed widespread variation between the treating hospitals (p = 0.001), and even between high-volume hepatopancreatobiliary hospitals (p = 0.0006). Younger age, positive lymph nodes, and positive surgical resection margins predicted an increased likelihood of receiving adjuvant treatment. Among patients receiving adjuvant treatment, positive margins and a low comorbidity burden were associated with crt compared with ctx.

CONCLUSIONS

Interinstitutional medical practice variation contributes significantly to differential patterns in the rate of adjuvant treatment for pcc. Whether such variation is warranted or unwarranted requires further investigation.

Does a strategy to promote shared decision-making reduce medical practice variation in the choice of either single or double embryo transfer after in vitro fertilisation? A secondary analysis of a randomised controlled trial

OBJECTIVES

The hypothesis that shared decision-making (SDM) reduces medical practice variations is increasingly common, but no evidence is available. We aimed to elaborate further on this, and to perform a first exploratory analysis to examine this hypothesis. This analysis, based on a limited data set, examined how SDM is associated with variation in the choice of single embryo transfer (SET) or double embryo transfer (DET) after in vitro fertilisation (IVF). We examined variation between and within hospitals.

DESIGN

A secondary analysis of a randomised controlled trial.

SETTING

5 hospitals in the Netherlands.

PARTICIPANTS

222 couples (woman aged <40 years) on a waiting list for a first IVF cycle, who could choose between SET and DET (ie, ≥2 embryos available).

INTERVENTION

SDM via a multifaceted strategy aimed to empower couples in deciding how many embryos should be transferred. The strategy consisted of decision aid, support of IVF nurse and the offer of reimbursement for an extra treatment cycle. Control group received standard IVF care.

OUTCOME MEASURE

Difference in variation due to SDM in the choice of SET or DET, both between and within hospitals.

RESULTS

There was large variation in the choice of SET or DET between hospitals in the control group. Lower variation between hospitals was observed in the group with SDM. Within most hospitals, variation in the choice of SET or DET appeared to increase due to SDM. Variation particularly increased in hospitals where mainly DET was chosen in the control group.

CONCLUSIONS

Although based on a limited data set, our study gives a first insight that including patients' preferences through SDM results in less variation between hospitals, and indicates another pattern of variation within hospitals. Variation that results from patient preferences could be potentially named the informed patient rate. Our results provide the starting point for further research.

TRIAL REGISTRATION NUMBER

NCT00315029; Post-results.

Inter-practice variation in polypharmacy prevalence amongst older patients in primary care

PURPOSE

Complex medication management in older people with multiple chronic conditions can introduce practice variation in polypharmacy prevalence. This study aimed to determine the inter-practice variation in polypharmacy prevalence and examine how this variation was influenced by patient and practice characteristics.

METHODS

This cohort study included 45,731 patients aged 55 years and older with at least one prescribed medication from 126 general practices that participated in NIVEL Primary Care Database in the Netherlands. Medication dispensing data of the year 2012 were used to determine polypharmacy. Polypharmacy was defined as the chronic and simultaneous use of at least five different medications. Multilevel logistic regression models were constructed to quantify the polypharmacy prevalence variation between practices. Patient characteristics (age, gender, socioeconomic status, number, and type of chronic conditions) and practice characteristics (practice location and practice population) were added to the models.

RESULTS

After accounting for differences in patient and practice characteristics, polypharmacy rates varied with a factor of 2.4 between practices (from 12.4% to 30.1%) and an overall mean of 19.8%. Age and type of conditions were highly positively associated with polypharmacy, and to a lesser extent a lower socioeconomic status.

CONCLUSIONS

Considerable variation in polypharmacy rates existed between general practices, even after accounting for patient and practice characteristics, which suggests that there is not much agreement concerning medication management in this complex patient group. Initiatives that could reduce inappropriate heterogeneity in medication management can add value to the care delivered to these patients. Copyright © 2016 John Wiley & Sons, Ltd.

Adoption of diagnostic technology and variation in caesarean section rates: a test of the practice style hypothesis in Norway

OBJECTIVE

To examine whether the introduction of advanced diagnostic technology in maternity care has led to less variation in type of delivery between hospitals in Norway.

DATA SOURCES

The Medical Birth Registry of Norway provided detailed medical information for 1.7 million deliveries from 1967 to 2005. Information about diagnostic technology was collected directly from the maternity units.

STUDY DESIGN

The data were analyzed using a two-level binary logistic model with Caesarean section as the outcome measure. Level one contained variables that characterized the health status of the mother and child. Hospitals are level two. A heterogeneous variance structure was specified for the hospital level, where the error variance was allowed to vary according to the following types of diagnostic technology: two-dimensional ultrasound, cardiotocography, ST waveform analysis, and fetal blood analyses.

PRINCIPAL FINDING

There was a marked variation in Caesarean section rates between hospitals up to 1973. After this the variation diminished markedly. This was due to the introduction of ultrasound and cardiotocography.

CONCLUSION

Diagnostic technology reduced clinical uncertainty about the diagnosis of risk factors of the mother and child during delivery, and variation in type of delivery between hospitals was reduced accordingly. The results support the practice style hypothesis.