Climate Classification is an Important Factor in Assessing Quality-of-Care Across Hospitals

Estimating the burden of disease attributable to high ambient temperature across climate zones: methodological framework with a case study

BACKGROUND

With high temperature becoming an increasing health risk due to a changing climate, it is important to quantify the scale of the problem. However, estimating the burden of disease (BoD) attributable to high temperature can be challenging due to differences in risk patterns across geographical regions and data accessibility issues.

METHODS

We present a methodological framework that uses Köppen-Geiger climate zones to refine exposure levels and quantifies the difference between the burden observed due to high temperatures and what would have been observed if the population had been exposed to the theoretical minimum risk exposure distribution (TMRED). Our proposed method aligned with the Australian Burden of Disease Study and included two parts: (i) estimation of the population attributable fractions (PAF); and then (ii) estimation of the BoD attributable to high temperature. We use suicide and self-inflicted injuries in Australia as an example, with most frequent temperatures (MFTs) as the minimum risk exposure threshold (TMRED).

RESULTS

Our proposed framework to estimate the attributable BoD accounts for the importance of geographical variations of risk estimates between climate zones, and can be modified and adapted to other diseases and contexts that may be affected by high temperatures.

CONCLUSIONS

As the heat-related BoD may continue to increase in the future, this method is useful in estimating burdens across climate zones. This work may have important implications for preventive health measures, by enhancing the reproducibility and transparency of BoD research.

Thirty-day hospital readmission prediction model based on common data model with weather and air quality data

Although several studies have attempted to develop a model for predicting 30-day re-hospitalization, few attempts have been made for sufficient verification and multi-center expansion for clinical use. In this study, we developed a model that predicts unplanned hospital readmission within 30 days of discharge; the model is based on a common data model and considers weather and air quality factors, and can be easily extended to multiple hospitals. We developed and compared four tree-based machine learning methods: decision tree, random forest, AdaBoost, and gradient boosting machine (GBM). Above all, GBM showed the highest AUC performance of 75.1 in the clinical model, while the clinical and W-score model showed the best performance of 73.9 for musculoskeletal diseases. Further, PM10, rainfall, and maximum temperature were the weather and air quality variables that most impacted the model. In addition, external validation has confirmed that the model based on weather and air quality factors has transportability to adapt to other hospital systems.

Development and evaluation of MADDIE: Method to Acquire Delivery Date Information from Electronic health records

OBJECTIVE

To develop an algorithm that infers patient delivery dates (PDDs) and delivery-specific details from Electronic Health Records (EHRs) with high accuracy; enabling pregnancy-level outcome studies in women's health.

MATERIALS AND METHODS

We obtained EHR data from 1,060,100 female patients treated at Penn Medicine hospitals or outpatient clinics between 2010-2017. We developed an algorithm called MADDIE: Method to Acquire Delivery Date Information from Electronic Health Records that infers a PDD for distinct deliveries based on EHR encounter dates assigned a delivery code, the frequency of code usage, and the time differential between code assignments. We validated MADDIE's PDDs against a birth log independently maintained by the Department of Obstetrics and Gynecology.

RESULTS

MADDIE identified 50,560 patients having 63,334 distinct deliveries. MADDIE was 98.6 % accurate (F₁-score 92.1 %) when compared to the birth log. The PDD was on average 0.68 days earlier than the true delivery date for patients with only one delivery (± 1.43 days) and 0.52 days earlier for patients with more than one delivery episode (± 1.11 days).

DISCUSSION

MADDIE is the first algorithm to successfully infer PDD information using only structured delivery codes and identify multiple deliveries per patient. MADDIE is also the first to validate the accuracy of the PDD using an external gold standard of known delivery dates as opposed to manual chart review of a sample.

CONCLUSION

MADDIE augments the EHR with delivery-specific details extracted with high accuracy and relies only on structured EHR elements while harnessing temporal information and the frequency of code usage to identify accurate PDDs.

Assessing the impact of drug courts on provider-directed marketing efforts by manufactures of medications for the treatment of opioid use disorder

BACKGROUND

Opioid use disorder (OUD) has become an increasingly consequential public health concern, especially in the United States where 47,600 opioid overdose deaths occurred in 2017 (Scholl, Seth, Kariisa, Wilson, & Baldwin, 2019). Medications for OUD (MOUD) are effective for decreasing opioid-related morbidity and mortality, including within the criminal justice system (Hedrich et al., 2012; Medications for Opioid Use Disorder Save Lives, 2019; Moore et al., 2019).While a stronger evidence base exists for agonist MOUD than for antagonist MOUD, a national study of drug courts found that half prohibited agonist MOUD (Matusow et al., 2013).Furthermore, recent media reports suggest that the pharmaceutical manufacturer of an antagonist MOUD has marketed its product towards drug court judges (Goodnough & Zernike, 2017; Harper, 2017). However, no study to date has systematically examined the relationship between MOUD marketing practices and drug courts. This ecological study examines the association at the county level between MOUD manufacturer payments to prescribers and drug court locations.

METHOD

We extracted provider-directed payments from Centers for Medicare and Medicaid Services (CMS)'s Sunshine Act Open Payments data 2014-2017, isolating those records mentioning any MOUD. We compared provider-directed payments for two major MOUDs: buprenorphine and extended-release naltrexone, in counties with and without drug courts.

RESULTS

The presence of any adult drug courts in the county is associated with a 7.86 percentage-point increase in the likelihood of providers in that county receiving any MOUD-related payments (about 22.46% of the sample mean, p<0.001) and with a 10.70% increase in the amount of these payments per 1000 county residents (p<0.001). The association between other forms of drug courts such as juvenile drug courts and Driving-Under-the-Influence courts (DUI) courts are less significant and slightly smaller in magnitude compared to those of adult drug courts. We did not find significant difference between payments by the manufacturer of Vivitrol and manufacturers of Zubsolv, Bunavail, and Suboxone (oral forms of buprenorphine).

CONCLUSIONS

Our results show an ecological association at the county level between MOUD manufacturer payments to prescribers and drug court presence. However, we did not examine a causal association between these variables.

Comparison of Rural vs Urban Direct-to-Physician Commercial Promotion of Medications for Treating Opioid Use Disorder

IMPORTANCE

In the United States, access to medications prescribed for opioid use disorder (OUD) is lower in rural counties than in urban counties. Considering the positive associations between direct-to-physician promotion of opiates and OUD medications and their prescribing rates, a study examining the association between pharmaceutical promotion of these medications and county-level rurality has merit.

OBJECTIVE

To assess whether rural counties received less pharmaceutical promotion of OUD medications compared with urban counties.

DESIGN, SETTING, AND PARTICIPANTS

This cross-sectional county-level study used all reported direct-to-physician pharmaceutical payments from manufacturers of medications prescribed for OUD from January 1, 2014, through December 31, 2017, as well as demographic and economic data at the county level from 3140 US counties. Logistic regression was used with year and state-level fixed effects to compare rural county and urban county odds of receiving any promotion of OUD medications. A negative binomial model was used with year and state-level fixed effects to compare the mean pharmaceutical payments per physician and per population in rural vs urban counties.

MAIN OUTCOMES AND MEASURES

A binary indicator for whether physicians in a county received any promotion related to OUD medications in a year. The second outcome was the value of promotion (eg, meals), with dollar amount of payments for each county by year. Counties were separated into metropolitan, micropolitan, and rural categories using the National Center for Health Statistics Urban-Rural Classification Scheme.

RESULTS

Of 3140 US counties with 18 318 physicians to whom promotion of OUD medications was directed, 1166 (37.1%) were metropolitan (16 740 physicians [91.4%]), 641 (20.4%) were micropolitan (1049 physicians [5.7%]), and 1333 (42.5%) were rural (529 physicians [2.9%]). Compared with physicians in metropolitan counties, physicians in rural counties had reduced odds of receiving any promotion (adjusted odds ratio, 0.57; 95% CI, 0.44-0.74) and received lower payments (adjusted incidence rate ratio, 0.24; 95% CI, 0.17-0.34).

CONCLUSIONS AND RELEVANCE

The study findings suggest that promotion for OUD medications is less likely to occur in rural counties and that this difference in promotion of OUD medications may be associated with differential commercial costs and benefits of promotion in rural settings.

Provider-directed marketing may increase prescribing of medications for opioid use disorder

BACKGROUND

Opioid use disorder (OUD) has become an increasingly grave public health concern, especially in the United States where approximately 80% of the global opioid supply is consumed. Despite greater awareness of the present overdose crisis, potentially life-saving OUD pharmacotherapy (medications for opioid use disorder or MOUD) utilization remains low. This study examines the extent of provider-directed marketing (detailing) for MOUD drugs and identifies any associations between a provider's receipt of detailing and their prescribing of MOUD drugs to Medicare Part D beneficiaries.

METHOD

We combined Open Payments data on all provider-directed payments from pharmaceutical manufacturers with physician-level data on all MOUD prescriptions filled in Medicare Part D. We estimated the adjusted difference in Medicare days supply for all MOUD drugs (collectively) and separately for each MOUD drug that was associated with receipt of payments.

RESULTS

The Open Payments data show that $7.0 million MOUD-specific promotional payments were made by pharmaceutical manufacturers to 12,056 US physicians from 2014 to 2016, which is <1/6th of the $50.3 million made in overall non-MOUD opioid-related promotional payments to 76,992 US physicians during that same period. Prescribers who received any MOUD-specific payments prescribed 1080 daily MOUD-related doses per year more than peers who did not receive any MOUD-specific payments (p < 0.001). The data also show the relatively greater association between receipt of detailing and Suboxone prescriptions compared to Vivitrol.

CONCLUSIONS

Provider-directed marketing by MOUD manufacturers has been found to be significantly and positively associated with incidence of MOUD prescribing in Medicare Part D, as well as with the quantity of MOUD prescribed.

Uncovering exposures responsible for birth season - disease effects: a global study

Objective

Birth month and climate impact lifetime disease risk, while the underlying exposures remain largely elusive. We seek to uncover distal risk factors underlying these relationships by probing the relationship between global exposure variance and disease risk variance by birth season.

Material and Methods

This study utilizes electronic health record data from 6 sites representing 10.5 million individuals in 3 countries (United States, South Korea, and Taiwan). We obtained birth month–disease risk curves from each site in a case-control manner. Next, we correlated each birth month–disease risk curve with each exposure. A meta-analysis was then performed of correlations across sites. This allowed us to identify the most significant birth month–exposure relationships supported by all 6 sites while adjusting for multiplicity. We also successfully distinguish relative age effects (a cultural effect) from environmental exposures.

Results

Attention deficit hyperactivity disorder was the only identified relative age association. Our methods identified several culprit exposures that correspond well with the literature in the field. These include a link between first-trimester exposure to carbon monoxide and increased risk of depressive disorder (R = 0.725, confidence interval [95% CI], 0.529-0.847), first-trimester exposure to fine air particulates and increased risk of atrial fibrillation (R = 0.564, 95% CI, 0.363-0.715), and decreased exposure to sunlight during the third trimester and increased risk of type 2 diabetes mellitus (R = −0.816, 95% CI, −0.5767, −0.929).

Conclusion

A global study of birth month–disease relationships reveals distal risk factors involved in causal biological pathways that underlie them.