Interrupted time series design is the strongest, quasi-experimental approach for evaluating longitudinal effects of interventions. Segmented regression analysis is a powerful statistical method for estimating intervention effects in interrupted time series studies. In this paper, we show how segmented regression analysis can be used to evaluate policy and educational interventions intended to improve the quality of medication use and/or contain costs.

Abstractobjectives To investigate antibiotic use in five national household surveys conducted with the WHO methodology to identify key determinants of antibiotic use in the community.methods Data from The Gambia, Ghana, Kenya, Nigeria and Uganda surveys were combined. We used logistic regression models that accounted for the clustered survey design to identify the determinants of care seeking outside the home and antibiotic use for 2914 cases of recent acute illness.results Overall, 95% of individuals with acute illness took medicines, 90% sought care outside their homes and 36% took antibiotics. In multivariate analyses, illness severity was a strong predictor of seeking care outside the home. Among those who sought outside care, the strongest predictor of antibiotic use was the presence of upper respiratory symptoms (OR: 3.02, CI: 2.36-3.86, P < 0.001), followed by gastrointestinal symptoms or difficulty breathing, and antibiotics use was less likely if they had fever. The odds of receiving antibiotics were higher when visiting a public hospital or more than one healthcare facility.conclusions The nature and severity of symptoms and patterns of care seeking had the greatest influence on decisions to take antibiotics. Antibiotics were widely available and inappropriately used in all settings. Policies to regulate antibiotics distribution as well as interventions to educate prescribers, dispensers and consumers are needed to improve antibiotic use.keywords antibiotic resistance, household surveys, appropriate use of medicines, Africa

Research Objective Patient cost‐sharing has been increasing over the past two decades, in particular due to high‐deductible health plans (HDHP). Asthma is a very common, serious, chronic disease in the United States. We analyzed a large, national sample of asthma patients to provide new evidence on OOP spending overall, across types of asthma care, and variation by patient income. Study Design We used 11 years (2004‐2014) of enrollment, claims, and geocoded census tract data on income. Annual OOP costs were measured as the sum of actual patient‐paid deductible, coinsurance and copayments on all medical care; spending was adjusted to 2014 dollars using the Medical CPI. Annual asthma OOP spending included OOP payments for asthma services (ie, asthma medications, office and emergency department (ED) visits, and hospitalizations with ICD‐9 codes for asthma, spacers for inhalers, asthma nebulizers). Patient income was measured using the census tract income associated with the patient’s address, and then, patients were categorized into income quintiles based on the distribution of household income across the United States (eg, not within sample). We report the association of OOP spending with patient income based on linear regression models controlling for patient age, sex, state, morbidity and asthma severity (using the Johns Hopkins ACG system), and year; models were stratified by annual HDHP enrollment. Population Studied All patients ages 4‐64 years with asthma (defined as having an asthma ICD‐9 diagnosis code for an outpatient or ED visit or hospitalization) in a large, national Commercial and Medicare Advantage claims database (n = 2 018 178). Principal Findings More patients in the lowest‐income quintile had a HDHP. Within plan type, levels of spending associated with income quintiles were similar but represented a significantly larger proportion of income for the lowest‐income quintile (Table). Conclusions Because patients spent similar amounts of OOP for asthma care, the lowest‐income quintile patients experienced greater cost burden. The majority of OOP asthma spending was on asthma medications, but lowest‐income quintile patients had greater spending on high acuity care than those in higher income quintiles. Implications for Policy or Practice Efforts should be made to understand whether these differences in cost burden are associated with cost‐related underuse of medications or other adverse outcomes to inform policy and insurance benefit design. Primary Funding Source Patient‐Centered Outcomes Research Institute.