Background: Individuals from minoritized racial and ethnic backgrounds experience pernicious and pervasive health disparities that have emerged, in part, from clinician bias. Objective: We used a natural language processing approach to examine whether linguistic markers in electronic health record (EHR) notes differ based on the race and ethnicity of the patient. To validate this methodological approach, we also assessed the extent to which clinicians perceive linguistic markers to be indicative of bias. Methods: In this cross-sectional study, we extracted EHR notes for patients who were aged 18 years or older; had more than 5 years of diabetes diagnosis codes; and received care between 2006 and 2014 from family physicians, general internists, or endocrinologists practicing in an urban, academic network of clinics. The race and ethnicity of patients were defined as White non-Hispanic, Black non-Hispanic, or Hispanic or Latino. We hypothesized that Sentiment Analysis and Social Cognition Engine (SEANCE) components (ie, negative adjectives, positive adjectives, joy words, fear and disgust words, politics words, respect words, trust verbs, and well-being words) and mean word count would be indicators of bias if racial differences emerged. We performed linear mixed effects analyses to examine the relationship between the outcomes of interest (the SEANCE components and word count) and patient race and ethnicity, controlling for patient age. To validate this approach, we asked clinicians to indicate the extent to which they thought variation in the use of SEANCE language domains for different racial and ethnic groups was reflective of bias in EHR notes. Results: We examined EHR notes (n=12,905) of Black non-Hispanic, White non-Hispanic, and Hispanic or Latino patients (n=1562), who were seen by 281 physicians. A total of 27 clinicians participated in the validation study. In terms of bias, participants rated negative adjectives as 8.63 (SD 2.06), fear and disgust words as 8.11 (SD 2.15), and positive adjectives as 7.93 (SD 2.46) on a scale of 1 to 10, with 10 being extremely indicative of bias. Notes for Black non-Hispanic patients contained significantly more negative adjectives (coefficient 0.07, SE 0.02) and significantly more fear and disgust words (coefficient 0.007, SE 0.002) than those for White non-Hispanic patients. The notes for Hispanic or Latino patients included significantly fewer positive adjectives (coefficient −0.02, SE 0.007), trust verbs (coefficient −0.009, SE 0.004), and joy words (coefficient −0.03, SE 0.01) than those for White non-Hispanic patients. Conclusions: This approach may enable physicians and researchers to identify and mitigate bias in medical interactions, with the goal of reducing health disparities stemming from bias.

The Peer Recovery Expansion Project (PREP) was designed to expand outreach and deliver enhanced treatment services via peer-recovery coaches for individuals with substance use disorder (SUD) and li...

Objective: To validate the use of electronic algorithms based on International Classification of Diseases (ICD)-10 codes to identify outpatient visits for urinary tract infections (UTI), one of the most common reasons for antibiotic prescriptions. Methods: ICD-10 symptom codes (e.g., dysuria) alone or in addition to UTI diagnosis codes plus prescription of a UTI-relevant antibiotic were used to identify outpatient UTI visits. Chart review (gold standard) was performed by two reviewers to confirm diagnosis of UTI. The positive predictive value (PPV) that the visit was for UTI (based on chart review) was calculated for three different ICD-10 code algorithms using (1) symptoms only, (2) diagnosis only, or (3) both. Results: Of the 1087 visits analyzed, symptom codes only had the lowest PPV for UTI (PPV = 55.4%; 95%CI: 49.3–61.5%). Diagnosis codes alone resulted in a PPV of 85% (PPV = 84.9%; 95%CI: 81.1–88.2%). The highest PPV was obtained by using both symptom and diagnosis codes together to identify visits with UTI (PPV = 96.3%; 95%CI: 94.5–97.9%). Conclusions: ICD-10 diagnosis codes with or without symptom codes reliably identify UTI visits; symptom codes alone are not reliable. ICD-10 based algorithms are a valid method to study UTIs in primary care settings.

Background: SARS CoV-2 virus (COVID-19) impacted the practice of healthcare in the United States, with technology being used to facilitate access to care and reduce iatrogenic spread. Since then, patient message volume to primary care providers has increased. However, the volume and trend of electronic communications after lockdown remain poorly described in the literature. Methods: All incoming inbox items (telephone calls, refill requests, and electronic messages) sent to providers from patients amongst four primary care clinics were collected. Inbox item rates were calculated as a ratio of items per patient encountered each week. Trends in inbox rates were assessed during 12 months before and after lockdown (March 1st, 2020). Logistic regression was utilized to examine the effects of the lockdown on inbox item rate post-COVID-19 lockdown as compared to the pre-lockdown period. Results: Before COVID-19 lockdown, 2.07 new inbox items per encounter were received, which increased to 2.83 items after lockdown. However, only patient-initiated electronic messages increased after lockdown and stabilized at a rate higher than the pre-COVID-19 period (aRR 1.27, p-value < 0.001). In contrast, prescription refill requests and telephone calls quickly spiked, then returned to pre-lockdown levels. Conclusion: Based on our observations, providers experienced a quick increase in all inbox items. However, only electronic messages had a sustained increase, exacerbating the workload of administrators, staff, and clinical providers. This study directly correlates healthcare technology adoption to a significant disruptive event but also shows additional challenges to the healthcare system that must be considered with these changes.

Background: Identifying underlying commonalities among all-cause encephalitis cases can be extraordinarily useful in predicting meaningful risk factors associated with inpatient mortality. Methods: A retrospective cohort of patients with encephalitis was derived from a clinical chart review of adult patients (age ≥18 years) across 16 different hospitals in Houston, Texas, between January 2005 and July 2015. Clinical features at admission were assessed for their correlation with inpatient mortality and used to derive a final risk score prediction tool. Results: The study included a total of 273 adult patients with all-cause encephalitis, 27 (9.9%) of whom died during hospitalization. A limited number of clinical features were substantially different between patients who survived and those who died (Charlson score, Glasgow coma scale [GCS], immunosuppression, fever on admission, multiple serologic studies, and abnormal imaging). A final multivariable logistic model was derived with the following risk factors, which were transformed into a scoring system: 1 point was assigned to the presence of a Charlson score >2, thrombocytopenia, or cerebral edema, and 2 points for a GCS value <8. Patients were then classified into different risk groups for inpatient mortality: 0 points (0%), 1 point (7%), 2 points (10.9%), 3 points (36.8%), and ≥4 points (81.8%). Conclusions: The risk score developed from this study shows a high predictive value. This can be highly beneficial in alerting care providers to key clinical risk factors associated with in-hospital mortality in adults with encephalitis.

West Nile virus (WNV) is a widespread and devastating disease, especially in those who develop neuroinvasive disease. A growing body of evidence describes sequelae years after infection, including neurological complications and chronic kidney disease (CKD). Eighty-nine out of 373 WNV-positive cases were followed for approximately two years and compared to 127 WNV-negative controls with and without CKD. Adjusted risk ratios (aRRs) were calculated via a log binomial regression to determine the impact of WNV exposure and other possible confounders on the likelihood of developing CKD. Cytokine profiles of WNV patients and controls were evaluated to characterize differences and describe potential underlying pathophysiological mechanisms. The associated risk for developing CKD was significantly associated with history of WNV infection (aRR = 1.91, 95% CI 1.13–3.25). Additionally, five distinct cytokines were found to be significantly associated with WNV infection (eotaxin, IL-8, IL-12p70, IP-10, and TNFα) after the p-value was adjusted to <0.0019 due to the Bonferroni correction. These data support that WNV infection is an independent risk factor for CKD, even after accounting for confounding comorbidities. WNV participants who developed CKD had high activity of proinflammatory markers, indicating underlying inflammatory disease. This study provides new insights into CKD resultant of WNV infection.