A predictive algorithm to identify ever smoking in medical claims-based epidemiologic studies.

PURPOSE: To develop and validate an algorithm to estimate probability of ever smoking using administrative claims. METHODS: Using population-based samples of Medicare-aged individuals (121,278 Behavioral Risk Factor Surveillance System survey respondents and 207,885 Medicare beneficiaries), we developed a logistic regression model to predict probability of ever smoking from demographic and claims data. We applied the model in 1,657,266 additional Medicare beneficiaries and calculated area under the receiver operating characteristic curve (AUC) using presence or absence of a tobacco-specific diagnosis or procedure code as our "gold standard." We used these "gold standard" and lung/laryngeal cancer codes to over-ride predicted probability as 100%. We calculated Spearman's rho between probability from this full algorithm and smoking assessed in prior Parkinson disease studies, by substituting our observed and prior ("true") smoking-Parkinson disease odds ratios into the attenuation equation. RESULTS: The predictive model contained 23 variables, including basic demographics, high alcohol consumption, asthma, cardiovascular disease and associated risk factors, selected cancers, and indicators of routine medical usage. The AUC was 67.6% (95% confidence interval 67.5%-67.7%) comparing smoking probability to tobacco-specific diagnosis or procedure codes. Spearman's rho for the full algorithm was 0.82. CONCLUSIONS: Ever smoking might be approximated in administrative data for use as a continuous, probabilistic variable in epidemiologic analyses.

Biologic targets of prescription medications and risk of neurodegenerative disease in United States Medicare beneficiaries.

OBJECTIVE: To identify prescription medications associated with a lower risk of three neurodegenerative diseases: Parkinson disease, Alzheimer disease, and amyotrophic lateral sclerosis. METHODS: We conducted a population-based, case-control study of U.S. Medicare beneficiaries in 2009 (42,885 incident neurodegenerative disease cases, 334,387 randomly selected controls). Using medication data from 2006-2007, we categorized all filled medications according to their biological targets and mechanisms of action on those targets. We used multinomial logistic regression models, while accounting for demographics, indicators of smoking, and health care utilization, to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for 141 target-action pairs and each neurodegenerative disease. For target-action pairs inversely associated with all three diseases, we attempted replication in a cohort study that included an active comparator group. We constructed the cohort by following controls forward for incident neurodegenerative disease from the beginning of 2010 until death or end of 2014, i.e., up to five years after the two-year exposure lag. We used Cox proportional hazards regression while accounting for the same covariates. RESULTS: The most consistent inverse association across both studies and all three neurodegenerative diseases was for xanthine dehydrogenase/oxidase blockers, represented by the gout medication, allopurinol. Allopurinol was associated with a 13-34% lower risk for each neurodegenerative disease group in multinomial regression, and a mean reduction of 23% overall, as compared to individuals who did not use allopurinol. In the replication cohort we observed a significant 23% reduction for neurodegenerative disease in the fifth year of follow-up, when comparing allopurinol users to non-users, and more marked associations with an active comparator group. We observed parallel associations for a related target-action pair unique to carvedilol. DISCUSSION/CONCLUSION: Xanthine dehydrogenase/oxidase blockade might reduce risk of neurodegenerative disease. However, further research will be necessary to confirm that the associations related to this pathway are causal or to examine whether this mechanism slows progression.

Incidence of amyotrophic lateral sclerosis in older adults.

INTRODUCTION/AIMS: We investigated the age- and sex-specific incidence and survival of Medicare beneficiaries with amyotrophic lateral sclerosis (ALS) in patients 66 to 90 years of age. METHODS: We identified all incident ALS cases within a population-based sample of Medicare beneficiaries in 2009 (total: 22 000 177 person-years at risk for ALS). We calculated age- and sex-specific incidence in 2009 according to multiple, progressively more stringent case definitions. Our most inclusive definition required one ALS code, whereas the most restrictive definition required at least one additional ALS code more than 6 months after the first code, including one from a neurologist. We identified associated imaging studies and electrodiagnostic testing and followed all cases through the end of 2014 to determine survival. RESULTS: The overall incidence for our most inclusive definition was 22.84 per 100 000 person-years for men and 16.05 per 100 000 person-years for women. The overall incidence was 5.72 per 100 000 person-years for men and 3.99 per 100 000 person-years for women for our most restrictive definition. For our most inclusive definition, fewer than 39.7% of cases ever had an ALS diagnosis from a neurologist, more than 50% had an electrodiagnostic test or imaging study, and 40.1% survived less than 1 year after diagnosis, with 25.5% of these cases surviving no more than 6 months. Cases not meeting the most restrictive definition were more likely than those who did meet the restrictive definition to be older, black, or Asian. DISCUSSION: The oldest and marginalized Medicare beneficiaries diagnosed with ALS are less likely to be included in epidemiological studies with restrictive definitions, but future studies will need to assess the accuracy of diagnosis.

A comparison of prediction approaches for identifying prodromal Parkinson disease.

Identifying people with Parkinson disease during the prodromal period, including via algorithms in administrative claims data, is an important research and clinical priority. We sought to improve upon an existing penalized logistic regression model, based on diagnosis and procedure codes, by adding prescription medication data or using machine learning. Using Medicare Part D beneficiaries age 66-90 from a population-based case-control study of incident Parkinson disease, we fit a penalized logistic regression both with and without Part D data. We also built a predictive algorithm using a random forest classifier for comparison. In a combined approach, we introduced the probability of Parkinson disease from the random forest, as a predictor in the penalized regression model. We calculated the receiver operator characteristic area under the curve (AUC) for each model. All models performed well, with AUCs ranging from 0.824 (simplest model) to 0.835 (combined approach). We conclude that medication data and random forests improve Parkinson disease prediction, but are not essential.

Fractures in the prodromal period of Parkinson disease.

OBJECTIVE: To examine the association between fractures and Parkinson disease (PD) during the 5-year prodromal phase as compared to controls. METHODS: We performed a population-based case-control study of Medicare beneficiaries in the United States from 2004 to 2009. We identified 89,632 incident PD cases and 117,760 comparable controls 66-90 years of age in 2009. PD case status was the outcome, and noncranial fracture the independent variable. We used logistic regression models to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for association between fracture and PD in yearly time intervals prior to PD diagnosis/control reference date, after adjusting for covariates. RESULTS: There were 39,606 total fractures (25.4% cases, 14.3% controls) over the 5 years prior to the PD diagnosis/control reference date. PD was positively associated with fractures even after adjusting for age, sex, race/ethnicity, Charlson comorbidity index, alcohol use, tobacco use, and osteoporosis. The association between PD and fracture was evident at yearly time windows prior to PD diagnosis/control reference date. The association between PD and each type of fracture strengthened as the PD diagnosis/control reference date approached (all time interaction p values ≤0.02). Among beneficiaries with a mechanism of injury, the majority were attributed to falls (74.6% cases, 72.8% controls). CONCLUSION: Fractures occur more commonly during the prodromal period of PD compared to controls, especially as diagnosis date approached, suggesting that patients with PD may experience unrecognized motor and nonmotor symptoms.

Herpesvirus Infections and Risk of Parkinson's Disease.

INTRODUCTION: Herpesviruses might play a role in the pathogenesis of neurodegenerative disorders. We sought to examine a possible association between alpha herpesvirus infections and Parkinson's disease. METHODS: We conducted a population-based case-control study of incident Parkinson's disease in 2009 Medicare beneficiaries age 66-90 years (89,790 cases, 118,095 randomly selected comparable controls). We classified beneficiaries with any diagnosis code for "herpes simplex" and/or "herpes zoster" in the previous 5 years as having had the respective alpha herpesviruses. In beneficiaries with Part D prescription coverage, we also identified those prescribed anti-herpetic medications. We calculated odds ratios (OR) and 95% CI between alpha herpesvirus diagnosis/treatment and Parkinson's disease with logistic regression, with adjustment for age, sex, race/ethnicity, smoking, and use of medical care. RESULTS: Parkinson's disease risk was inversely associated with herpes simplex (OR 0.79, 95% CI 0.74-0.84), herpes zoster (OR 0.88, 95% CI 0.85-0.91), and anti-herpetic medications (OR 0.87, 95% CI 0.80-0.96). CONCLUSION: Herpesvirus infection or treatment might reduce risk of Parkinson's disease, but future studies will be required to explore whether this inverse association is causal.

Use of medical care biases associations between Parkinson disease and other medical conditions.

OBJECTIVE: To examine how use of medical care biases the well-established associations between Parkinson disease (PD) and smoking, smoking-related cancers, and selected positively associated comorbidities. METHODS: We conducted a population-based, case-control study of 89,790 incident PD cases and 118,095 randomly selected controls, all Medicare beneficiaries aged 66 to 90 years. We ascertained PD and other medical conditions using ICD-9-CM codes from comprehensive claims data for the 5 years before PD diagnosis/reference. We used logistic regression to estimate age-, sex-, and race-adjusted odds ratios (ORs) between PD and each other medical condition of interest. We then examined the effect of also adjusting for selected geographic- or individual-level indicators of use of care. RESULTS: Models without adjustment for use of care and those that adjusted for geographic-level indicators produced similar ORs. However, adjustment for individual-level indicators consistently decreased ORs: Relative to ORs without adjustment for use of care, all ORs were between 8% and 58% lower, depending on the medical condition and the individual-level indicator of use of care added to the model. ORs decreased regardless of whether the established association is known to be positive or inverse. Most notably, smoking and smoking-related cancers were positively associated with PD without adjustment for use of care, but appropriately became inversely associated with PD with adjustment for use of care. CONCLUSION: Use of care should be considered when evaluating associations between PD and other medical conditions to ensure that positive associations are not attributable to bias and that inverse associations are not masked.

Inflammatory bowel disease and risk of Parkinson's disease in Medicare beneficiaries.

INTRODUCTION: Gastrointestinal (GI) dysfunction precedes the motor symptoms of Parkinson's disease (PD) by several years. PD patients have abnormal aggregation of intestinal α-synuclein, the accumulation of which may be promoted by inflammation. The relationship between intestinal α-synuclein aggregates and central nervous system neuropathology is unknown. Recently, we observed a possible inverse association between inflammatory bowel disease (IBD) and PD as part of a predictive model of PD. Therefore, the objective of this study was to examine the relationship between PD risk and IBD and IBD-associated conditions and treatment. METHODS: Using a case-control design, we identified 89,790 newly diagnosed PD cases and 118,095 population-based controls >65 years of age using comprehensive Medicare data from 2004-2009 including detailed claims data. We classified IBD using International Classification of Diseases version 9 (ICD-9) diagnosis codes. We used logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs) to evaluate the association between PD and IBD. Covariates included age, sex, race/ethnicity, smoking, Elixhauser comorbidities, and health care use. RESULTS: PD was inversely associated with IBD overall (OR = 0.85, 95% CI 0.80-0.91) and with both Crohn's disease (OR = 0.83, 95% CI 0.74-0.93) and ulcerative colitis (OR = 0.88, 95% CI 0.82-0.96). Among beneficiaries with ≥2 ICD-9 codes for IBD, there was an inverse dose-response association between number of IBD ICD-9 codes, as a potential proxy for IBD severity, and PD (p-for-trend = 0.006). CONCLUSION: IBD is associated with a lower risk of developing PD.

Traumatic brain injury in the prodromal period of Parkinson's disease: A large epidemiological study using medicare data.

OBJECTIVE: Studies suggest a greater risk of Parkinson's disease (PD) after traumatic brain injury (TBI), but it is possible that the risk of TBI is greater in the prodromal period of PD. We aimed to examine the time-to-TBI in PD patients in their prodromal period compared to population-based controls. METHODS: We identified 89,790 incident PD cases and 118,095 comparable controls aged > 65 years in 2009 using Medicare claims data. Using data from the preceding 5 years, we compared time-to-TBI in PD patients in their prodromal period to controls. We estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for TBI in a Cox regression, while adjusting for age, sex, race/ethnicity, modified Charlson comorbidity index, smoking, and alcohol use. RESULTS: Risk of TBI was greater in PD patients in their prodromal period across all age and sex groups, with HRs consistently increasing with proximity to PD diagnosis. HRs ranged from 1.64 (95% CI, 1.52, 1.77) 5 years preceding diagnosis to 3.93 (95% CI, 3.74, 4.13) in the year before. The interaction between PD, TBI, and time was primarily observed for TBI attributed to falls. Motor dysfunction and cognitive impairment, suggested by corresponding International Classification of Diseases, Ninth Revision codes, partially mediated the PD-TBI association. INTERPRETATION: There is a strong association between PD and a recent TBI in the prodromal period of PD. This association strengthens as PD diagnosis approaches and may be a result of undetected nonmotor and motor symptoms, but confirmation will be required. Ann Neurol 2017;82:744-754.

A predictive model to identify Parkinson disease from administrative claims data.

OBJECTIVE: To use administrative medical claims data to identify patients with incident Parkinson disease (PD) prior to diagnosis. METHODS: Using a population-based case-control study of incident PD in 2009 among Medicare beneficiaries aged 66-90 years (89,790 cases, 118,095 controls) and the elastic net algorithm, we developed a cross-validated model for predicting PD using only demographic data and 2004-2009 Medicare claims data. We then compared this model to more basic models containing only demographic data and diagnosis codes for constipation, taste/smell disturbance, and REM sleep behavior disorder, using each model's receiver operator characteristic area under the curve (AUC). RESULTS: We observed all established associations between PD and age, sex, race/ethnicity, tobacco smoking, and the above medical conditions. A model with those predictors had an AUC of only 0.670 (95% confidence interval [CI] 0.668-0.673). In contrast, the AUC for a predictive model with 536 diagnosis and procedure codes was 0.857 (95% CI 0.855-0.859). At the optimal cut point, sensitivity was 73.5% and specificity was 83.2%. CONCLUSIONS: Using only demographic data and selected diagnosis and procedure codes readily available in administrative claims data, it is possible to identify individuals with a high probability of eventually being diagnosed with PD.