Multiple Comorbidities, Psychiatric Disorders, Healthcare Resource Utilization and Costs Among Adults with Essential Tremor: A Retrospective Observational Study in a Large US Commercially Insured and Medicare Advantage Population

Background: Essential tremor (ET), the most common movement disorder, often impairs patients’ ability to perform activities of daily living, mental health, and quality of life. Objectives: To assess comorbidities, psychiatric disorders, healthcare resource utilization (HCRU), and costs among patients with ET compared with patients without ET. Methods: This retrospective observational study was conducted using a large US administrative claims database. Patients with ET were identified during the study period (1/1/2017–12/31/2019). The earliest claim date with ET diagnosis was identified as the index date. An index date was assigned randomly for each non-ET patient. Patients had to be at least 22 years old and be enrolled in the health plan for at least 6 months before and at least 12 months after the index date. Patients with and those without ET were matched 1:1 on age, gender, payer type, and first 3 digits of their ZIP code. Comorbidities were assessed using data within 6 months prior to the index date. Psychiatric disorders, HCRU, and costs were examined using data within 12 months after the index date. Results: The mean (SD) age of ET patients (n = 5286) was 70.8 (11.8) years, 49.1% were female, and 82.9% were Medicare Advantage members. In the 12 months following the index date, 26.0% of patients had no insurance claims for ET-related pharmacotherapy or invasive therapies. Patients with ET had a higher number of comorbidities than non-ET patients (5.3 [3.2] vs 4.0 [3.3]); a higher prevalence of psychiatric disorders (depression: 25.6% vs 15.3%; adjusted odds ratio (AOR) [95% CI], 1.56 [1.41-1.73]; anxiety: 27.7% vs 15.5%, AOR: 1.78 [1.61-1.96]); and higher total healthcare costs: $17 560 [$39 972] vs $13 237 [$27 098], adjusted cost ratio [95% CI]: 1.11 [1.06-1.16]; all P<.0001. Discussion: Highly prevalent multiple comorbidities and psychiatric disorders should be considered in the context of clinical decision-making to optimize ET management. Conclusions: This study represents the largest observational study to report ET disease and economic burdens in a real-world setting. The data demonstrate increased comorbidity, mental health, and healthcare cost burdens among ET patients compared with matched non-ET patients. These findings underscore the need for innovative care for this complex population.


INTRODUCTION
Essential tremor (ET) is one of the most common neurological movement disorders in adults. 1,2 ET prevalence is around 4% in persons age 40 and older and increases considerably with age, affecting an estimated 20% of individuals in their 90s and over worldwide. 3 A new metaanalysis demonstrated that the pooled prevalence for all ages was 1.33% worldwide, and gender did not impact the prevalence of ET. 4 Louis et al 5 reported that ET affects approximately 2.2% of the US population, or 7 million individuals in the United States. Recently published analysis of claims data from 819 661 patients in the United States found that the ET prevalence rate increased at an annual growth rate of 6.27% from 2010 to 2018, with 7.29% and 2.05% annual growth rates for the Medicare Advantage and commercially insured populations, respectively. 6 JOURNAL OF HEALTH ECONOMICS AND OUTCOMES RESEARCH While traditionally regarded as benign, ET is in fact a chronic, progressive, and disabling neurological disease that causes uncontrolled rhythmic shaking, most often affecting the hands. [7][8][9] It is characterized by the presence of an action tremor as well as a myriad of symptoms such as gait disturbances, postural instability, and cognitive impairment. 7,9 These symptoms may worsen over time and significantly impact a patient's quality of life and ability to perform activities of daily living. 2,[10][11][12] Medications used to treat ET can be moderately effective at reducing symptoms but do not treat the underlying disease. 13,14 It is estimated that only 30% to 50% of patients with ET respond to pharmacotherapy. 11,15 Invasive therapies include deep brain stimulation (DBS), magnetic resonance-guided focused ultrasound (MRgFUS), and thalamotomy for unilateral and DBS for bilateral procedures. 13,16 Only 3% of patients with ET whose tremors are refractory to pharmacotherapy choose to undergo DBS. 17 Both thalamotomy and DBS have been shown to be highly effective at reducing limb tremor magnitude; however, they are invasive procedures and associated with significant risks of side effects. 18,19 A newly developed surgical alternative, MRgFUS, may be ideal for patients with substantial comorbidities but can be applied to only a very limited number of patients with ET due to the selective criteria. In addition, because it involves ablation of brain tissue, MRgFUS may result in permanent side effects. 13,20 Recent research has explored novel noninvasive, nonpharmacological treatment, such as noninvasive transcutaneous patterned afferent stimulation to aid in the symptomatic relief of hand tremor; these therapies may mitigate some of the existing treatment gaps. 2,16,[21][22][23] There is limited real-world evidence on the burden of comorbidities, mental disorders, healthcare resource utilization (HCRU), and healthcare costs (HCC) among patients with ET. Such evidence is important for clinical and policy decision-making to optimize ET care and improve outcomes. This study aimed to assess the comorbidity burden, prevalence of psychiatric disorders, HCRU, and HCCs among adult patients with ET in the commercially insured and Medicare Advantage populations using a large US administrative claims database.

Study Design and Data Source
A retrospective observational cohort study of adult patients with ET was conducted using a large US administrative claims data from January 1, 2017, to December 31, 2019 (study period). The claims database included over 20 million medical insurance beneficiaries and contained patient enrollment data, as well as inpatient and outpatient medical and pharmacy claims for fully insured commercial health plan and Medicare Advantage members. 24 All data handling complied with federal and state requirements; the privacy and security of individually identifiable personal health information, required by the Health Insurance Portability and Accountability Act (HIPAA), were preserved. The study was approved by an independent institutional review board prior to initiation.

Patient Selection
Patients with ET were identified between July 1, 2017, and December 31, 2018 (index period). Patients were considered to have ET if they had at least 1 medical claim with an ET diagnosis code (International Classification of Diseases, Tenth Revision, Clinical Modification [ICD-10-CM] code of G25.0 in any position). The first observed ET diagnosis date was defined as the index date. An index date was assigned randomly for each patient without ET in the database. All patients were at least 22 years of age at the index date and had fully insured commercial health and/or Medicare Advantage coverage with medical and pharmacy health insurance benefits for at least 6 months prior to the index date (baseline period) and 12 months after the index date (follow-up period). Patients with a diagnosis of Parkinson's disease (ICD-10-CM: G20.x) or thyroid disorders (ICD-10-CM: E00.x-E03.x, E06.5, E07.9, E89.0, or thyroid hormone prescription) were excluded from the ET cohort. A non-ET comparison cohort was created using 1:1 exact matching on age, gender, payer type, first 3-digits of ZIP code, and index month.

Demographic Characteristics
Patient demographic characteristics included age at index date, gender, geographic region (Midwest, Northeast, South, and West), rural or urban residence, and median household income. Household income was estimated by merging 2010 census data to the claims data using ZIP code.

Assessment of Comorbidities
The 47 most common chronic conditions in the Aetna administrative claims data were identified using previously described methods based on ICD-10-CM codes (Online Supplemental Material, Table S1). 24,25 We also calculated Charlson Comorbidity Index (CCI) and ageadjusted CCI, 2 validated metrics that summarize disease burden and predict mortality and high HCC risk. [26][27][28][29] All comorbid conditions were assessed using all claims data within 6 months prior to the index date.

Assessment of Psychiatric Disorders
Psychiatric disorders were identified using ICD-10-CM codes (Online Supplemental Material, Table S2) during the 12-month follow-up period.

ET Treatment
Patients with ET were categorized into 1 of 3 treatment groups consistent with evidence of billed treatments in the claims database during the 12-month follow-up period: (1) pharmacotherapy, those patients given evidence of filled prescription for propranolol and other β-blockers, primidone, anti-epileptic drugs, benzodiazepines, or botulinum neurotoxin; (2) invasive therapy, those given evidence of DBS, thalamotomy, focused ultrasound, radio-surgical gamma knife thalamotomy; or (3) untreated, those who did not receive any ETrelated pharmacotherapy or invasive therapy.

Healthcare Resource Utilization and Costs
Healthcare resource utilization included any emergency department (ED) visits, inpatient admission, and length of hospital stays (LOS) among those with inpatient admissions. All-cause HCRU was assessed using all medical claims; ET-related HCRU was assessed using all medical claims with a primary diagnosis of ET (ICD-10-CM: G25.0). Allcause HCC included costs of all medical claims and pharmacy claims; ET-related HCC included costs of medical claims with a primary diagnosis of ET and ET-related pharmacotherapy costs of pharmacy claims. HCRU and HCC were aggregated over the 12-month follow-up period. All HCC were inflated to 2019 US dollars using the medical care services component of the Consumer Price Index. 30

Statistical Analysis
Demographic and clinical characteristics and comorbidities were analyzed descriptively. Means (±SD) or medians (interquartile range) were reported for continuous variables, and frequencies (%) were reported for categorical variables. To compare differences between treatment groups, age groups (<65, ≥65 years old), and ET and non-ET groups, statistical significance was assessed with the Student t test, Wilcoxon rank-sum test, or Kruskal-Wallis test for continuous variables, and the Pearson χ 2 test for categorical variables. To evaluate the association of patient characteristics and the HCC, generalized linear models were performed to calculate adjusted cost ratios with the corresponding 95% confidence interval (CI). Log-transformation and gamma distribution were applied based on the distribution and presence of heteroskedasticity. 24,31 To evaluate the association of ET and the prevalence of psychiatric disorders and HCC, generalized linear models were performed to calculate adjusted odds ratios and adjusted cost ratio with the corresponding 95% CI, respectively. All data management and statistical analyses were conducted using SAS version 9.4 statistical software (SAS Institute Inc, Cary, North Carolina). All P values were 2-sided, with P < .05 considered statistically significant.

Patient Characteristics
Of the 5286 eligible patients with ET (Online Supplemental Material, Figure S1), 49.1% were female; median age was 72 years, with 79% over the age of 65; 37.5% were in the southern United States; and 50.4% lived in rural areas ( Table 1). Medicare Advantage patients made up 82.9% of the study population; the remaining 17.1% were commercially insured patients. Among patients with ET, 71.3% received some pharmacotherapy (pharmacotherapy group), 26.0% received no ET-related treatment (untreated group), and 2.7% had inva-

Comorbidity and Multimorbidity
The 5 most common comorbidities were hypertension (67.9% of patients with ET), pain disorders (61.8%), hyperlipidemia (55.4%), fatigue and sleep-related disorders (28.1%), and diabetes mellitus (27.3%) (see Table 2 for the complete list of top 30 most common comorbidities in the study population). The distribution of number of comorbid conditions in patients with ET varied by treatment type and age (Figure 1) Table S5). Multivariable analysis indicated age (P = .0032), region, living in an urban area (P < .05), higher household income (P < .05), commercial insurance (P < .0001), a higher number of comorbidities (P < .0001), presence of psychiatric disorders (P < .0001), and being in the invasive therapy group (P < .0001) were associated with increased all-cause HCCs. Similarly, older age, living in a suburban area, higher household income, a higher number of comorbidities, presence of psychiatric disorders, and being in the invasive therapy group were associated with increased ET-related HCCs (Supplemental Figures S4C  and S4D; Supplemental Table S5).   32 proposed that chronic stress, including posttraumatic stress disorder, depression, and anxiety, may not only be responsible for the comorbidities associated with ET but in some cases may directly and indirectly induce ET. The rates of depression found among patients with ET in this study were lower than the rates reported in other studies. In this analysis, 25.6% of patients were diagnosed with depression, whereas 2 separate articles reported much higher rates. 33,34 In one cross-sectional analysis in Spain, the self-reported depression rate was 43.8% in 235 patients with ET. 33 Another cross-sectional study, including 245 Han Chinese patients with ET, assessed depression using the Hamilton Depression Rating Scale-24 Item and showed that 54.3% had at least mild depression. 34 One possible explanation is that depression and other comorbid conditions such as fatigue, stress, and anxiety may not always cause the patient to seek care from a physician. Also, the studies from Spain and China included a small sample size, and the country, geographic re-gion, race/ethnicity, and other factors could lead to the variation in the depression rate. Our study demonstrated that 95.5% of patients with ET had at least 1 comorbidity, 79.7% had at least 3 comorbidities, and 56.9%    38 longer LOS, worse postoperative complications, and higher HCC and mortality. 24,[39][40][41] The interplay between multimorbidity and chronic disease is extremely important but receives inadequate attention in many chronic conditions. 24,[41][42][43] Data regarding the multiple comorbidities, their interactions, and the impacts on mental health and health economics among patients with ET are still scant. Comorbidities increase the complexity of disease management and pose a significant clinical and public health challenge. 24,[39][40][41][42][43][44][45] Multiple comorbidities may lead to escalating use of multiple medications (polypharmacy), which may increase the risk of inappropriate medi-cation use, accidental overdosing, poor medication or therapy adherence and persistence, adverse drug-drug interactions, and increased HCC. 24,[43][44][45] Pharmacotherapy is currently the mainstay for ET treatment. 13,14,46 However, it has been reported that the first-line agents are ineffective in 30% to 70% of patients due to lack of response, adverse effects, and poor adherence and persistence. 11,[13][14][15]46 Our data show the wide spectrum of comorbidities among patients with ET, supporting the need for noninvasive, nonpharmacological treatments to minimize risk of adverse events. Novel treatment options can empower healthcare providers and ET care teams to provide safe and effective therapies that can be tailored to the specific healthcare needs of each patient. Because conventional clinical trials of devices, surgery, and medications in ET often exclude patients with some comorbidities, 20-23 our data underscore the need for more inclusive pragmatic trials of all interventions in ET in the real-world setting. This study indicates that patients with ET had higher HCC than the patients without ET: average all-cause HCC was $17 560 PPPY, ET-related HCC accounted for 13% of total HCC, and ET-related treatment significantly increased total HCC. Patient demographic and clinical factors associated with both increased all-cause HCC and ET-related HCC included number of comorbidities, treatment type, psychiatric disorders, age, geographic region, rural or urban residence. In addition, commercial health insurance was associated with increased all-cause HCC, and higher household income was associated with increased ET-related HCC. Invasive therapy, multiple comorbidities, and presence of psychiatric disorders remain the major drivers of HCC among patients with ET. Future studies are needed to focus on cost-effectiveness of novel therapies and new approaches in ET management in both commercially insured and Medicare Advantage populations.

Limitations
There are several limitations in the study. First, this retrospective observational study using administrative healthcare claims data has inherent limitations because the data were collected for provider reimbursement and not for research purposes; therefore, certain comorbid conditions and outcomes may be underestimated. Second, since all comorbidities were identified based on ICD-10-CM codes alone, coding errors or misclassification are likely. However, this approach is commonly used in many other comorbidity studies. [24][25][26][27][28][29] Third, there are no ICD-10-CM codes specific to the degree of ET severity or tremor location. Fourth, the study results may not be generalizable to the overall population, as those with other types of health insurance or uninsured may have different characteristics.

CONCLUSIONS
While ET is traditionally regarded as benign, this study clearly highlights the significant impact of ET on patients, HCRU, and HCC. Multiple comorbidities and psychiatric disorders are highly prevalent among patients with ET and should be considered in the context of healthcare policy and medical decision-making to optimize ET care. Patients with ET are associated with substantial HCRU, posing a substantial economic burden to afflicted patients and the healthcare system. Invasive therapy, comorbidity burden, and presence of psychiatric disorders associated with ET remain major drivers for HCC. The data demonstrate increased mental health and HCC burden among patients with ET compared with matched patients without ET and underscore the need for innovation in care for this complex population.