Machine learning to identify chronic cough from administrative claims data.

Accurate identification of patient populations is an essential component of clinical research, especially for medical conditions such as chronic cough that are inconsistently defined and diagnosed. We aimed to develop and compare machine learning models to identify chronic cough from medical and pharmacy claims data. In this retrospective observational study, we compared 3 machine learning algorithms based on XG Boost, logistic regression, and neural network approaches using a large claims and electronic health record database. Of the 327,423 patients who met the study criteria, 4,818 had chronic cough based on linked claims-electronic health record data. The XG Boost model showed the best performance, achieving a Receiver-Operator Characteristic Area Under the Curve (ROC-AUC) of 0.916. We selected a cutoff that favors a high positive predictive value (PPV) to minimize false positives, resulting in a sensitivity, specificity, PPV, and negative predictive value of 18.0%, 99.6%, 38.7%, and 98.8%, respectively on the held-out testing set (n = 82,262). Logistic regression and neural network models achieved slightly lower ROC-AUCs of 0.907 and 0.838, respectively. The XG Boost and logistic regression models maintained their robust performance in subgroups of individuals with higher rates of chronic cough. Machine learning algorithms are one way of identifying conditions that are not coded in medical records, and can help identify individuals with chronic cough from claims data with a high degree of classification value.

Risk of chemotherapy-induced febrile neutropenia in intermediate-risk regimens: Clinical and economic outcomes of granulocyte colony-stimulating factor prophylaxis.

BACKGROUND: Chemotherapy-induced neutropenia increases the risk of febrile neutropenia (FN) and infection with resultant hospitalizations, with substantial health care resource utilization (HCRU) and costs. Granulocyte-colony stimulating factor (GCSF) is recommended as primary prophylaxis for chemotherapy regimens having more than a 20% risk of FN. Yet, for intermediate-risk (10%-20%) regimens, it should be considered only for patients with 1 or more clinical risk factors (RFs) for FN. It is unclear whether FN prophylaxis for intermediate-risk patients is being optimally implemented. OBJECTIVE: To examine RFs, prophylaxis use, HCRU, and costs associated with incident FN during chemotherapy. METHODS: This retrospective study used administrative claims data for commercial and Medicare Advantage enrollees with nonmyeloid cancer treated with intermediate-risk chemotherapy regimens during January 1, 2009, to March 31, 2020. Clinical RFs, GCSF prophylaxis, incident FN, HCRU, and costs were analyzed descriptively by receipt of primary GCSF, secondary GCSF, or no GCSF prophylaxis. Multivariable Cox regression analysis was used to examine the association between number of RFs and cumulative FN risk. RESULTS: The sample comprised 13,937 patients (mean age 67 years, 55% female). Patients had a mean of 2.3 RFs, the most common being recent surgery, were aged 65 years or greater, and had baseline liver or renal dysfunction; 98% had 1 or more RFs. However, only 35% of patients received primary prophylaxis; 12% received secondary prophylaxis. The hazard ratio of incident FN was higher with increasing number of RFs during the first line of therapy, yet more than 54% of patients received no prophylaxis, regardless of RFs. Use of GCSF prophylaxis varied more by chemotherapeutic regimen than by number of RFs. Among patients treated with rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride (doxorubicin hydrochloride), vincristine, and prednisone, 76% received primary prophylaxis, whereas only 22% of patients treated with carboplatin/paclitaxel received primary prophylaxis. Among patients with a first line of therapy FN event, 78% had an inpatient stay and 42% had an emergency visit. During cycle 1, mean FN-related coordination of benefits-adjusted medical costs per patient per month ($13,886 for patients with primary prophylaxis and $18,233 for those with none) were driven by inpatient hospitalizations, at 91% and 97%, respectively. CONCLUSIONS: Incident FN occurred more often with increasing numbers of RFs, but GCSF prophylaxis use did not rise correspondingly. Variation in prophylaxis use was greater based on regimen than RF number. Lower health care costs were observed among patients with primary prophylaxis use. Improved individual risk identification for intermediate-risk regimens and appropriate prophylaxis may decrease FN events toward the goal of better clinical and health care cost outcomes. DISCLOSURES: This work was funded by Sandoz Inc., which participated in the design of the study, interpretation of the data, writing and revision of the manuscript, and the decision to submit the manuscript for publication. The study was performed by Optum under contract with Sandoz Inc. The author(s) meet criteria for authorship as recommended by the International Committee of Medical Journal Editors. The authors received no direct compensation related to the development of the manuscript. Dr Li is an employee of Sandoz Inc. Drs Bell and Lal and Mr Peterson-Brandt were employees of Optum at the time of the study. Ms Anderson and Dr Aslam are employees of Optum. Dr Lyman has been primary investigator on a research grant from Amgen to their institution and has consulted for Sandoz, G1 Therapeutics, Partners Healthcare, BeyondSpring, ER Squibb, Merck, Jazz Pharm, Kallyope, Teva; Fresenius Kabi, Seattle Genetics, and Samsung.

Development of a natural language processing algorithm to detect chronic cough in electronic health records.

BACKGROUND: Chronic cough (CC) is difficult to identify in electronic health records (EHRs) due to the lack of specific diagnostic codes. We developed a natural language processing (NLP) model to identify cough in free-text provider notes in EHRs from multiple health care providers with the objective of using the model in a rules-based CC algorithm to identify individuals with CC from EHRs and to describe the demographic and clinical characteristics of individuals with CC. METHODS: This was a retrospective observational study of enrollees in Optum's Integrated Clinical + Claims Database. Participants were 18-85 years of age with medical and pharmacy health insurance coverage between January 2016 and March 2017. A labeled reference standard data set was constructed by manually annotating 1000 randomly selected provider notes from the EHRs of enrollees with ≥ 1 cough mention. An NLP model was developed to extract positive or negated cough contexts. NLP, cough diagnosis and medications identified cough encounters. Patients with ≥ 3 encounters spanning at least 56 days within 120 days were defined as having CC. RESULTS: The positive predictive value and sensitivity of the NLP algorithm were 0.96 and 0.68, respectively, for positive cough contexts, and 0.96 and 0.84, respectively, for negated cough contexts. Among the 4818 individuals identified as having CC, 37% were identified using NLP-identified cough mentions in provider notes alone, 16% by diagnosis codes and/or written medication orders, and 47% through a combination of provider notes and diagnosis codes/medications. Chronic cough patients were, on average, 61.0 years and 67.0% were female. The most prevalent comorbidities were respiratory infections (75%) and other lower respiratory disease (82%). CONCLUSIONS: Our EHR-based algorithm integrating NLP methodology with structured fields was able to identify a CC population. Machine learning based approaches can therefore aid in patient selection for future CC research studies.

Real-world characteristics, treatment experiences and corticosteroid utilisation of patients treated with tofacitinib for moderate to severe ulcerative colitis.

BACKGROUND: Tofacitinib is an oral, small molecule JAK inhibitor for the treatment of UC. We aimed to describe the real-world treatment experience and corticosteroid utilisation of patients treated with tofacitinib in a US claims database. METHODS: Patients with a UC diagnosis who initiated tofacitinib, vedolizumab or tumour necrosis factor inhibitor (TNFi) treatment between May 2018 and July 2019 were identified from the Optum Research Database. Demographic and clinical characteristics of patients who initiated tofacitinib, vedolizumab or TNFi were described. Oral corticosteroid use prior to and following tofacitinib initiation was evaluated. Tofacitinib adherence (proportion of days covered) and continuation was assessed for 6 months following initiation. Analyses were descriptive and stratified by prior biologic use (naïve, 1 or ≥ 2; minimum of 12 months prior to tofacitinib initiation). RESULTS: Among patients initiating tofacitinib (N = 225), mean age was 45.6 (SD 16.5) years and 50.2% were female. Of these, 43 (19.1%) patients were biologic-naïve and 182 (80.9%) had prior biologic use (92 [40.9%], 1 prior biologic; 90 [40.0%], ≥ 2 prior biologics). Among patients with 1 prior biologic, 82.6% were previously treated with a TNFi. Among patients with ≥ 2 prior biologics, 54.4% were previously treated with vedolizumab and a TNFi, 16.7% with two TNFi and 28.9% with ≥ 3 prior biologics. In the 6 months prior to tofacitinib initiation, 65.8% of patients had received oral corticosteroids (74.4%, 60.9% and 66.7% for biologic-naïve, 1 and ≥ 2 prior biologics, respectively). The proportion of patients with ongoing oral corticosteroid use 3-6 months after tofacitinib initiation decreased to 13.3% (9.3%, 18.5% and 10.0% for biologic-naïve, 1 and ≥ 2 prior biologics, respectively), and 19.6% of patients discontinued oral corticosteroid use during the 6 months after tofacitinib initiation. Overall, tofacitinib adherence, as determined by the mean proportion of days covered during the 6-month follow-up, was 0.7 (median 0.8). During the 6-month follow-up, 84.9% of patients continued tofacitinib. CONCLUSIONS: Among patients with UC initiating tofacitinib, the majority had prior biologic use. Tofacitinib adherence was high, discontinuation was low and oral corticosteroid utilisation decreased irrespective of prior biologic use. Further research with longer follow-up and a larger sample size is required.

Treatment Patterns, Depression, and Anxiety Among US Patients Diagnosed with Hyperhidrosis: A Retrospective Cohort Study.

INTRODUCTION: Hyperhidrosis is associated with social and emotional stress due to limitations on health-related quality of life. This study examined real-world treatment patterns and concomitant depression and/or anxiety in patients with hyperhidrosis. METHODS: Commercial health plan members in the US with ≥ 2 hyperhidrosis diagnosis codes and/or antiperspirant prescription claims were identified from January 2010 through November 2017. A control cohort (CC) of patients without hyperhidrosis was matched to the hyperhidrosis cohort on demographic characteristics. Depression and/or anxiety were identified by ≥ 1 relevant diagnosis code or pharmacy claim. A multivariable logistic regression model estimated odds of treatment in the hyperhidrosis cohort, and depression/anxiety in the hyperhidrosis cohort and CC, adjusting for patient characteristics. RESULTS: A total of 44,484 patients with hyperhidrosis were identified, of whom 58.5% were female, with a mean (± standard deviation) age of 36.5 ± 16.5 years (83.5% ≥ 18 years). A small majority of patients (51.6%, 0.69/person-year) received treatment with prescription antiperspirants. Post-index oral systemic therapies, medical procedures, and surgical options were uncommon. At 12 months post-index, 48.4% of the sample had not filled a prescription for extra- or prescription-strength antiperspirants. Compared with the CC (n = 137,451), a higher percentage of patients with hyperhidrosis had depression or anxiety reported during follow-up (41.1 vs. 28.2%, p < 0.001); this corresponded to higher adjusted odds of depression/anxiety in patients with hyperhidrosis [odds ratio (OR) 1.76, 95% confidence interval (CI) 1.72-1.80, p < 0.001]. Baseline depression and/or anxiety were associated with lower odds of receiving hyperhidrosis treatment (OR 0.77, 95% CI 0.73-0.80), as was increasing age and male gender. Patients with hyperhidrosis also had more frequent incident depression/anxiety during follow-up (18.2 vs. 10.6%, p < 0.001). CONCLUSION: In this real-world analysis, hyperhidrosis was associated with increased odds of depression and/or anxiety. However, relatively low percentages of patients received prescription topical or oral treatments or underwent surgery, suggesting that tolerability, efficacy, and provider awareness may be limiting factors in the effective treatment of hyperhidrosis.

Real-world cost of treatment for multiple sclerosis patients initiating and receiving infused disease-modifying therapies per recommended label in the United States.

Aims: The study evaluated the real-world cost of treatment in multiple sclerosis (MS) patients initiating infused disease-modifying-therapies (DMT) in the United States.Materials and Methods: This retrospective cohort study using administrative claims data included adult patients with MS initiating index infusion DMT (ocrelizumab (OCR), natalizumab (NTZ) or alemtuzumab (ATZ)) from April 2017-September 2018 with 6-months pre/12-months post-index continuous enrollment. The primary cohort included patients who had prescribed annual dosing visits indicated by the approved product label (PL): 3 OCR, 5 ATZ, and 12-13 NTZ infusion visits within the first year of initiation. Annual treatment cost was the sum of all costs on index DMT infusion visit dates. Costs were summarized for a primary and secondary cohort of patients receiving additional doses than prescribed in PL (>3 OCR, >5 ATZ, and >13 NTZ infusion visits); and an overall cohort of patients who met minimum required annual dose (≥3 OCR, ≥5 ATZ, and ≥12 NTZ), further stratified by insurance type.Results: For patients in the primary cohort (123 OCR, 18 ATZ, and 48 NTZ), mean (standard-deviation) annual cost of treatment with OCR, ATZ, and NTZ cohorts was $72,066 ($34,480), $121,053 ($51,097) and $93,777 ($38,815), respectively. Among patients initiating OCR and NTZ, 15 and 6% respectively, had additional infusion visits leading to greater costs. Mean annual costs of index infusion DMT treatment in the overall cohort (162 patients treated with OCR, 18 with ATZ, 56 with NTZ) were $80,582, $121,053, and $93,807, respectively. The mean costs for commercial enrollees were higher than those for MAPD enrollees.Limitations: Small sample size, limited population generalizability, and cost-reduction for ATZ beyond the second year need to be accounted for.Conclusions: Real-world infusion DMT treatment costs for commercially insured patients were higher than perceived expenditures based on wholesale acquisition cost and administration costs via a physician-fee schedule. Consideration of real-world costs in cost-effectiveness and treatment/coverage decisions is needed.