Enhancing case ascertainment of Parkinson's disease using Medicare claims data in a population-based cohort: the Cardiovascular Health Study

Completeness of cohort-linked U.S. Medicare data: An example from the Agricultural Health Study (1999–2016)

•

We describe linked U.S. Medicare claims data in the Agricultural Health Study.

•

Incomplete claims data were related to geographic, demographic, and health.

•

We saw potential informative missingness by pesticide use and mortality.

•

Incomplete data in Medicare-linked cohorts may impact sample size and validity.

Medicare Fee for Service (FFS) claims data, including inpatient (Part A) and outpatient (Part B) services, provide a valuable resource for research on older adults (≥65 year) in linked U.S. cohorts. Here we describe our experience linking the Agricultural Health Study cohort, including 47,501 licensed pesticide applicators and spouses from North Carolina (NC) and Iowa (IA) to Medicare claims data from 1999 to 2016. Given increased Part C (i.e., managed care/Medicare Advantage) enrollment during this period, and a resulting lack of available Part C claims data prior to 2015, we also explored potential for informative missingness. We compared those with partial or limited/no FFS to those with complete FFS coverage (i.e., ≥11 months per year parts AB, but not C, throughout Medicare enrollment) in relation to baseline farm size, general pesticide use, and mortality, in logistic regression models adjusted for age, sex, race, education, and smoking, and stratified by state. While 46,689 participants (98%) were linked to Medicare IDs, only 33,487 (70%) had complete FFS, 9353 (20%) had partial FFS (≥1 year FFS but not complete), and 3849 (8%) had limited/no FFS (Part A or Part C-only). Incomplete FFS was more common in NC, mostly due to Part C, and was associated with farm characteristics, pesticide use, and mortality. These findings indicate that, in addition to reduced sample size in analyses limited to complete FFS, missingness may not be random. The potential impact of incomplete FFS data and changes in coverage type need to be considered when planning linked analyses and interpreting results.

Parkinson's disease case ascertainment in a large prospective cohort

Background

In epidemiologic studies where physician-based case adjudication is not feasible, Parkinson’s disease (PD) case ascertainment is often limited to self-reports which may not be accurate. We evaluated strategies to identify PD cases in the Agricultural Health Study (AHS).

Methods

Doctor-diagnosed PD was self-reported on all cohort-wide surveys; potential cases were also identified from death certificates. Follow-up surveys asked about PD-related motor and non-motor symptoms. For PD confirmation, we collected additional diagnosis, symptom, and treatment data from 510 potential PD cases or their proxy (65% of those contacted) in a supplemental screener and obtained medical records for a subset (n = 65). We classified PD cases using established criteria and screener data.

Results

Of 510 potential PD cases, 75% were considered “probable” or “possible”; this proportion increased among participants diagnosed by a specialist (81.2%), taking PD medication (85.2%), or reporting ≥5 motor symptoms (86.8%) in a regular AHS survey. Of those with medical records, 93% (57 of 61) of probable or possible PD was confirmed. Never-smoking and non-motor and motor symptoms reported in prior AHS surveys were more common with probable/possible PD than unconfirmed PD.

Conclusion

In this retrospective PD case ascertainment effort, we found that PD self-report with information on motor symptoms or medications may be a reasonable alternative for identifying PD cases when physician exam is not feasible. Because of intervening mortality, screeners could not be obtained from about one-third of those contacted. Thus, findings warrant replication.

Parkinson's disease case ascertainment in prospective cohort studies through combining multiple health information resources

Epidemiological evidence from prospective cohort studies on risk factors of Parkinson's disease (PD) is limited as case ascertainment is challenging due to a lack of registries and the disease course of PD. The objective of this study was to create a case ascertainment method for PD within two prospective Dutch cohorts based on multiple sources of PD information. This method was validated using clinical records from the general practitioners (GPs). Face validity of the case ascertainment was tested for three etiological factors (smoking, sex and family history of PD). In total 54825 participants were included from the cohorts AMIGO and EPIC-NL. Sources of PD information included self-reported PD, self-reported PD medication, a 9 item screening questionnaire (Tanner), electronical medical records, hospital discharge data and mortality records. Based on these sources we developed a likelihood score with 4 categories (no PD, unlikely PD, possible PD, likely PD). For the different sources of PD information and for the likelihood score we present the agreement with GP-validated cases. Risk of PD for established factors was studied by logistic regression as exact diagnose dates were not always available. Based on the algorithm, we assigned 346 participants to the likely PD category. GP validation confirmed 67% of these participants in EPIC-NL, but only 12% in AMIGO. PD was confirmed in only 3% of the participants with a possible PD classification. PD case ascertainment by mortality records (91%), EMR ICPC (82%) and self-reported information (62-69%) had the highest confirmation rates. The Tanner PD screening questionnaire had a lower agreement (18%). Risk estimates for smoking, family history and sex using all likely PD cases were comparable to the literature for EPIC-NL, but not for smoking in AMIGO. Using multiple sources of PD evidence in cohorts remains important but challenging as performance of sources varied in validity.

Predictors of incident epilepsy in older adults: The Cardiovascular Health Study

OBJECTIVE

To determine the prevalence, incidence, and predictors of epilepsy among older adults in the Cardiovascular Health Study (CHS).

METHODS

We analyzed data prospectively collected in CHS and merged with data from outpatient Medicare administrative claims. We identified cases with epilepsy using self-report, antiepileptic medication, hospitalization discharge ICD-9 codes, and outpatient Medicare ICD-9 codes. We used Cox proportional hazards regression to identify factors independently associated with incident epilepsy.

RESULTS

At baseline, 42% of the 5,888 participants were men and 84% were white. At enrollment, 3.7% (215 of 5,888) met the criteria for prevalent epilepsy. During 14 years of follow-up totaling 48,651 person-years, 120 participants met the criteria for incident epilepsy, yielding an incidence rate of 2.47 per 1,000 person-years. The period prevalence of epilepsy by the end of follow-up was 5.7% (335 of 5,888). Epilepsy incidence rates were significantly higher among blacks than nonblacks: 4.44 vs 2.17 per 1,000 person-years (p < 0.001). In multivariable analyses, risk of incident epilepsy was significantly higher among blacks compared to nonblacks (hazard ratio [HR] 4.04, 95% confidence interval [CI] 1.99-8.17), those 75 to 79 compared to those 65 to 69 years of age (HR 2.07, 95% CI 1.21-3.55), and those with history of stroke (HR 3.49, 95% CI 1.37-8.88).

CONCLUSIONS

Epilepsy in older adults in the United States was common. Blacks, the very old, and those with history of stroke have a higher risk of incident epilepsy. The association with race remains unexplained.

Groundwater Pesticide Levels and the Association With Parkinson Disease

It is unclear whether exposure to environmentally relevant levels of pesticides in groundwater is associated with an increased risk of Parkinson disease (PD). The purpose of this study was to examine the relationship between PD and pesticide levels in groundwater. This cross-sectional study included 332 971 Medicare beneficiaries, including 4207 prevalent cases of PD from the 2007 Colorado Medicare Beneficiary Database. Residential pesticide levels were estimated from a spatial model based on 286 well water samples with atrazine, simazine, alachlor, and metolachlor measurements. A logistic regression model with known PD risk factors was used to assess the association between residential groundwater pesticide levels and prevalent PD. We found that for every 1.0 µg/L of pesticide in groundwater, the risk of PD increases by 3% (odds ratio = 1.03; 95% confidence interval: 1.02-1.04) while adjusting for age, race/ethnicity, and gender suggesting that higher age-standardized PD prevalence ratios are associated with increasing levels of pesticides in groundwater.

Validation of secondary data sources to identify Parkinson disease against clinical diagnostic criteria

Parkinson disease (PD) is the second most common neurodegenerative disorder. Its diagnosis relies solely on a clinical examination and is not straightforward because no diagnostic test exists. Large, population-based, prospective cohort studies designed to examine other outcomes that are more common than PD might provide cost-efficient alternatives for studying the disease. However, most cohort studies have not implemented rigorous systematic screening for PD. A majority of epidemiologic studies that utilize population-based prospective designs rely on secondary data sources to identify PD cases. Direct validation of these secondary sources against clinical diagnostic criteria is lacking. The Framingham Heart Study has prospectively screened and evaluated participants for PD based on clinical diagnostic criteria. We assessed the predictive value of secondary sources for PD identification relative to clinical diagnostic criteria in the Framingham Heart Study (2001-2012). We found positive predictive values of 1.0 (95% confidence interval: 0.868, 1.0), 1.0 (95% confidence interval: 0.839, 1.0), and 0.50 (95% confidence interval: 0.307, 0.694) for PD identified from self-report, use of antiparkinsonian medications, and Medicare claims, respectively. The negative predictive values were all higher than 0.99. Our results highlight the limitations of using only Medicare claims data and suggest that population-based cohorts may be utilized for the study of PD determined via self-report or medication inventories while preserving a high degree of confidence in the validity of PD case identification.