To estimate the mean annual cost incurred by the health care system for a patient with APDS in the US, we developed a burden-of-illness cost model reflecting the perspective of US commercial payers (Fig. 1). The model combined clinical expert survey data on the annual prevalence of APDS manifestations with the US costs data related to manifestations associated with APDS. The inputs included the results of a clinical expert survey on manifestations and the cost data. Only the US direct medical costs were included in the model. The patient population comprised adults and adolescents 12 years and older with a diagnosis of APDS. Patient characteristics in the model, such as weight and body surface area, were sourced from a randomized, placebo-controlled, phase 3 trial in APDS [9]. The model did not account for patient subgroups and only considered a 1-year time horizon. Given that the model calculates annual manifestation costs, it was not necessary to discount the model results.

Cost model. APDS activated phosphoinositide 3-kinase delta syndrome and CNS central nervous system

After review of the scientific literature and discussion with clinicians, a list of manifestations associated with APDS was compiled [10,11,12,13]. Annual rates of manifestations were not available in the published literature to inform the model, given the rarity of the disease. A comprehensive survey was conducted to gather information relating to the annual prevalence of manifestations associated with APDS as well as the frequency at which they were experienced in clinical practice each year.

The survey content was developed from data from APDS-specific literature [10,11,12,13]. For each question, the annual prevalence was defined as the percentage of patients who were likely to have experienced a manifestation each year. Survey respondents were asked to assume that the manifestations were being treated and to consider all patients with APDS, not just those who they have directly treated. For example, those who did not have firsthand experience with the manifestations identified by the publications were asked to consider the experiences of APDS experts with whom they interact to estimate prevalence. This request of clinicians was required due to the rareness of the disease.

Six US-based clinicians who had experience treating patients with APDS responded to the survey. Collectively, survey respondents had treated a total of 27 patients, with a range of 2–8 patients per respondent at the time of the survey. With only 351 unique patients described globally in the literature as of 2023 and considering the rarity of APDS, the patient ranges reported among respondents are meaningful and highlight their expertise in APDS [8]. Two means were calculated from the clinician survey responses: (1) a standard arithmetic mean and (2) a weighted mean. For the weighted mean, weights were based on the number of patients treated by the clinician; greater weight was assigned to survey responses from clinicians who treated more patients with APDS.

Approximately 150 manifestation costs were used as inputs in the model and were categorized as infection (e.g., bacilli Calmette-Guérin disease, cellulitis, chronic rhinosinusitis, dental/oral tissue abscesses, endocarditis, fungal infection, granulomatous lesions, meningitis, ocular infections, otitis media, otitis needing tympanostomy tubes, pneumonia, respiratory tract infections, septicemia, Epstein–Barr virus, cytomegalovirus, herpes simplex virus, varicella-zoster virus, tonsillitis, and tonsillitis with tonsillectomy), hematology (e.g., cytopenias, pancytopenia, factor XI deficiency, factor IX deficiency, hemolytic anemia, hepatomegaly, immune thrombocytopenia, lymphadenopathy, nodular lymphoid hyperplasia, spherocytosis, splenomegaly, and thrombotic thrombocytopenic purpura), malignancy (e.g., diffuse large B-cell lymphoma [DLBCL], Hodgkin lymphoma, mucosa-associated lymphoid tissue [MALT] lymphoma, marginal zone B-cell lymphoma, multiple lymphoma, and other non-lymphoma malignancies), central nervous system (e.g., neuropsychiatric disorders, developmental delays, failure to thrive, seizures, and brain biopsy), pulmonary (e.g., atelectasis, bronchiectasis, and interstitial lung disease), gastrointestinal (e.g., enteropathy, graft-vs-host disease, eosinophilic esophagitis/eosinophilic gastrointestinal disease, inflammatory bowel disease, nodular regenerative hyperplasia of the liver, pancreatic insufficiency, gastrointestinal surgery, liver biopsy, and liver transplant), dermatology (e.g., eczema and dermatitis), endocrine (e.g., adrenal insufficiency, diabetes, and hypothyroidism), and other (e.g., rheumatoid arthritis, allergy, asthma, and cardiomyopathy). Manifestation costs were sourced from the literature.

Patient monitoring costs related to disease surveillance and testing costs incurred by patients were sourced from Centers for Medicare & Medicaid Services (CMS). Supplementary Table 1 presents the cost per test, and the total cost per year as well as the testing frequencies assumed in the model. Some costs, such as chemistry panel laboratory test and computed tomography scans, were derived from the sum and the mean of multiple CMS costs codes, respectively (Supplementary Table 2). CMS costs and other Medicare costs were inflated to commercial prices using a factor of 134%, as per the Medicare Payment Advisory Commission (MedPAC) source [14].

All cost inputs were inflated using the Consumer Price Index to represent 2023 prices (US dollar) (Supplementary Table 3). The frequency of manifestations associated with APDS was based on the results of the clinician survey, using standard means of survey responses in the base case and weighted means in the scenario analysis. For manifestations that can recur within a year (e.g., pneumonia), clinicians were asked to report the number of times in a year a patient may experience the manifestation; the reported number of episodes was included in the cost calculation for that manifestation. The inputs derived from the manifestations survey were included in the calculations for the proportion of patients who experienced manifestations associated with APDS and the resource use associated with each manifestation.

An example sequence of the cost calculation for pneumonia is provided in Supplementary Fig. 1.

A base case analysis applying the standard mean and a scenarios analysis using weighted mean results from the survey (referred to hereafter as weighted mean scenario analysis) were performed. In the base case analysis, values derived from the survey results were used to calculate the standard arithmetic mean, while the minimum and maximum results were used to determine the lower and upper estimates, respectively. The output from each analysis was the annual cost of manifestations associated with APDS per patient.

This article is based on clinical expert survey data and publicly available cost information and does not contain any new studies involving human participants or animals conducted by any of the authors.