Comparing the risk of serious infections in patients with and without MS: A German claims data analysis.

Background: Research suggests that serious infections (SIs), comorbidities, and advanced disability represent key drivers of early death in people with Multiple Sclerosis (pwMS). Nevertheless, further research is warranted to better characterize and quantify the risk of SI among pwMS compared to the general population.

Methods: Our study consisted of a retrospective analysis of claims data provided by a German statutory health insurance fund, AOK PLUS, covering 3.4 million individuals in Saxony and Thuringia from 01/01/2015-31/12/2019. A propensity score (PS) matching method was used to compare the incidence of SIs among people with and without MS. PwMS were required to have ≥1 inpatient or ≥2 confirmed outpatient diagnoses of MS (ICD-10 G35) from a neurologist from 01/01/2016-31/12/2018, while people from the general population could not have any inpatient/outpatient codes for MS during the entire study period. The index date was defined as the first observed MS diagnosis or, in the case of the non-MS cohort, a randomly assigned date within the inclusion period. For both cohorts a PS was assigned, corresponding with their probabilistic likelihood of having MS based on observable factors including patient characteristics, comorbidities, medication use and other variables. People with and without MS were matched using a 1:1 nearest neighbor strategy. An exhaustive list of ICD-10 codes was created in association with 11 main SI categories. SIs were those recorded as the main diagnosis during an inpatient stay. ICD-10 codes from the 11 main categories were sorted into smaller classification units, used to distinguish between infections. A 60-day threshold for measuring new cases was defined to account for the potential risk of re-infection. Patients were observed until the end of the study period (31/12/2019) or death. Cumulative incidence, incidence rates (IRs) and IR ratios (IRRs) were reported during follow-up and at 1-, 2- and 3-years post-index.

Results: A total of 4250 and 2,098,626 patients were included in the unmatched cohorts of people with and without MS. Ultimately, one match was identified for all 4,250 pwMS, corresponding with a final population of 8,500 patients. On average, patients were 52.0/52.2 years in the matched MS/non-MS cohorts; the gender breakdown was 72% female. Overall, IRs of SIs per 100 patient years (PY) were higher in pwMS than in those without MS (1 year: 7.6 vs. 4.3; 2 years: 7.1 vs. 3.8; 3 years: 6.9 vs. 3.9). During follow-up, the most common infection types in pwMS were of a bacterial/parasitic origin (2.3 per 100 PY), followed by respiratory (2.0) and genitourinary (1.9) infections. Respiratory infections were most common in patients without MS (1.5 per 100 PY). Differences in the IRs of SIs were statistically significant (p<0.01) at each measurement window, with IRRs ranging from 1.7-1.9. PwMS had a higher risk of hospitalized genitourinary infections (IRR: 3.3-3.8) and bacterial/parasitic infections (2.0-2.3).

Conclusions: The incidence of SIs is much higher in pwMS, than comparators from the general population in Germany. Differences in hospitalized infection rates were largely driven by higher levels of bacterial/parasitic and genitourinary infections in the MS population.