Pain Phenotypes and Pain Multimorbidity Among Medicare Beneficiaries With Cerebral Palsy

This cohort study used Medicare fee-for-service research identifiable files from the 20% random sample from January 1, 2008, through December 31, 2020, to identify a cohort of patients with CP. We used data from the Master Beneficiary Summary File, Medicare Provider Analysis and Review, outpatient file, and carrier file to identify these patients. The University of Michigan institutional review board deemed this study exempt and waived the need for informed consent as data were retrospective and deidentified. The study followed the STROBE reporting guideline. Patients with CP aged 18 years or older at enrollment were identified based on ICD-9 and ICD-10 codes. We excluded patients enrolled in Medicare Advantage or Part C or whose coverage lapsed from January 1, 2016, through December 31, 2020. We identified pain diagnoses using ICD-10 codes 5 (eTable in Supplement 1 ). Counts and proportions of demographic information, including age, sex, race and ethnicity, census region, dual enrollment status, and CP subtype, were calculated for the entire cohort, along with prevalence of pain. Race and ethnicity data were from beneficiary claims and included to show that the sample is representative of the US population. All analyses were conducted between October 3, 2023, and April 24, 2024, using SAS, version 9.4 (SAS Institute Inc).

We identified 24 464 eligible adults with CP (mean [SD], age 48.9 [13.8] years, 46.7% female, 53.3% male) ( Table 1 ). A total of 21 767 patients (89.0%) had 1 or more documented pain diagnoses, 2697 (11.0%) had no pain diagnosis, 3708 (15.2%) had a single pain condition throughout the study period, 18 059 (73.8%) exhibited pain multimorbidity (ie, ≥2 diagnoses), and 7259 (29.7%) had pain extreme multimorbidity (ie, ≥5 diagnoses). All nonmutually exclusive pain conditions are listed in Table 2 . Per Centers for Medicare & Medicaid Services reporting rules, no specific data can be provided for any cell with fewer than 11 individuals. The distribution of patients by pain phenotype was 21 101 (86.3%) with any evidence of nociceptive pain, 11 213 (45.8%) with any evidence of nociplastic pain, and 4139 (16.9%) with any evidence of neuropathic pain. The distribution of patients across the cohort was 9499 (38.8%) with nociceptive pain only; 7538 (30.8%) with nociceptive and nociplastic pain; 3065 (12.5%) with neuropathic, nociceptive, and nociplastic pain; 999 (4.1%) with neuropathic and nociceptive pain; 591 (2.4%) with nociplastic pain only; 56 (0.2%) with neuropathic pain only; and 19 (0.1%) with neuropathic and nociplastic pain.

This cohort study is, to our knowledge, the largest to examine pain and its subtypes by topologic phenotype. Our finding of 89.0% pain prevalence underscores the importance of understanding and treating pain in adults with CP. Furthermore, 73.8% of the patients had more than 1 pain diagnosis, which increases the complexity of treatment decisions and highlights the necessity of understanding etiology to deliver effective treatment. Pain subtypes and combinations of subtypes were similar across all clinical presentations of CP. A limitation of this study is that sensitivity and specificity of using ICD codes for the identification of CP and CP subtypes are unknown. These findings provide crucial information for clinical and public health audiences, especially given the comparison to the general population, where chronic pain is estimated to be present in approximately 20% of adults. 6