Healthcare resource utilization and payer cost analysis of patients with prurigo nodularis

Dear Editor, Prurigo nodularis (PN) is a chronic dermatosis characterized by severely pruritic nodules on the trunk and extensor surfaces. In addition to its skin manifestations, PN is also associated with a variety of systemic conditions.1,2,3 Furthermore, many patients with PN are refractory to therapy with no US Food and Drug Administration‐approved agents available to date.4 Despite challenges in management and a significant burden of comorbidities in PN, there are limited studies examining healthcare utilization in this condition.5 Herein we use a medical claims database from a large, ambulatory cohort to describe real‐world healthcare utilization and economic burden in PN. Study approval was granted by the Johns Hopkins School of Medicine Institutional Review Board. Data from the IBM MarketScan Commercial Claims database between October 2015 and December 2016 were analysed.1 During this period, the MarketScan database contains claims for 24 049 563 enrolees in the US.1 The prevalence of PN in this population has previously been estimated as 72 per 100 000.1 Patients with PN aged 18–64 years were identified using ≥ two claims with the corresponding International Classification of Disease (ICD)‐10‐CM diagnosis code (L28·1), while three cohorts without PN were selected for comparison: age‐ and sex‐matched controls, atopic dermatitis (AD) (L20·9) and psoriasis (L40·9).4,6 Relative cost ratios were calculated using generalized linear models with log link and gamma distribution. Incidence rate ratios (IRR) of being seen by a provider were calculated using negative binomial regression. Models were adjusted for age and sex as potential confounders. Bonferroni correction for multiple comparisons was applied (with P < 0·001 considered statistically significant). Mean total healthcare spending per patient for the 7095 patients with PN in USD during the 15‐month study period was $8334 (median $2228). The highest mean costs during the study period were incurred in the outpatient setting ($5580), followed by inpatient ($1896), emergency department ($464) and pharmacy/lab ($394). Cost ratios comparing PN with matched controls were increased: 1·52 for total costs (P < 0·001), 2·70 for pharmacy/lab (P < 0·001) and 1·67 for outpatient costs (P < 0·001). Cost ratios comparing overall healthcare expenditures in PN to AD (1·10, P = 0·030) and PN to psoriasis (0·74, P = 0·001), respectively, did not reach statistical significance in adjusted models. In most settings except inpatient, psoriasis had increased healthcare costs compared with PN (cost ratios 0·65–0·79, P < 0·001) in adjusted models. The majority of patients with PN were seen by dermatology (85·9%) and internal/family medicine (69·8%). Patients with PN were more than 30 times as likely to be seen by dermatology compared with controls and more than five times as likely as AD or psoriasis, respectively (P < 0·001) (Table 1). However, patients with PN were less likely to see an internal/family medicine physician compared with AD (IRR 0·76, P < 0·001) and psoriasis (IRR 0·62, P < 0·001) in adjusted models. Patients with PN were more likely to have seen psychiatry compared with controls (IRR 2·03), AD (IRR 1·73) and psoriasis (IRR 1·45), respectively, and also more likely to have seen a pain specialist compared with AD (IRR 2·05) and psoriasis (IRR 1·36), respectively (P < 0·001). Relative incidence rates of being seen by provider type CI, confidence interval; IRR, incidence rate ratio aCalculated using negative binomial regression; bmatched in age and sex to patients with prurigo nodularis; cadjusted for age and sex. Bold indicates statistically significant P‐value < 0·001. Relative incidence rates of being seen by provider type CI, confidence interval; IRR, incidence rate ratio aCalculated using negative binomial regression; bmatched in age and sex to patients with prurigo nodularis; cadjusted for age and sex. Bold indicates statistically significant P‐value < 0·001. In summary, PN incurs higher healthcare costs compared with age‐ and sex‐matched controls although overall costs compared with AD and psoriasis, respectively, did not reach statistical significance. While patients with PN were more likely to see dermatology and psychiatry, they were less likely to see an internal/family medicine provider compared with patients with other inflammatory skin conditions (i.e. AD, psoriasis). Not only does this reiterate the significant burden of mental health conditions in PN,1,7 it also demonstrates that many patients with PN (30·2%) do not receive longitudinal primary care follow‐up. For these patients, facilitating connection can aid in comprehensive management of associated conditions including HIV, renal and cardiovascular disease.1 Furthermore, better control of systemic conditions may help with symptoms of PN in patients whose underlying illnesses play a contributory role. Study limitations include the retrospective study design, as well as use of claims data of individuals aged 18–64 years with healthcare insurance which precluded stratification of results by disease severity and may limit the generalizability of results to other populations. Selection bias may be present in the comparison groups, given their identification through ICD‐10 codes for unspecified AD and psoriasis, respectively. Multivariable models accounting for age and sex were created to minimize the influence of potential confounders, but residual confounding may remain. Because PN may be undiagnosed given lack of familiarity with the condition, it is likely that conservative estimates of PN's economic burden are reported here. Future work should explore generalizability of these results to patients with PN in other demographics (e.g. the uninsured population, other age groups). Funding sources: S.G.K. received grant funding from the Skin of Color Society and is a recipient of the Dermatology Foundation Medical Dermatology Career Development Award. Conflicts of interest: S.G.K. is an advisory board member for Menlo and Trevi Therapeutics and has received grant funding from Kiniksa Pharmaceuticals. None of the other authors have disclosures to report.