Health care utilizations and costs of Campylobacter enteritis in Germany: A claims data analysis

Objective The number of reported cases of Campylobacter enteritis (CE) remains on a high level in many parts of the world. The aim of this study was to analyze the health care utilizations and direct and indirect costs of CE and sequelae of patients insured by a large health insurance with 26 million members in Germany. Methods Claims data of insurants with at least one CE diagnosis in 2017 (n = 13,150) were provided, of which 9,945 were included in the analysis of health care utilizations and costs. If medical services were not diagnosis-linked, CE-associated costs were estimated in comparison to up to three healthy controls per CE patient. Indirect costs were calculated by multiplying the work incapacities by the average labor costs. Total costs of CE in Germany were extrapolated by including all officially reported CE cases in 2017 using Monte Carlo simulations. Results Insurants showed a lower rate of 56 CE diagnoses per 100,000 than German surveillance data for 2017, but with a similar age, gender and regional distribution. Of those CE cases, 6.3% developed post-infectious reactive arthritis, Guillain-Barré syndrome (GBS), inflammatory bowel disease (IBD) and/or irritable bowel syndrome (IBS). Health care utilizations differed depending on CE severity, age and gender. Average CE-specific costs per patient receiving outpatient care were € 524 (95% CI 495–560) over a 12-month period, whereas costs per hospitalized CE case amounted to € 2,830 (2,769–2,905). The analyzed partial costs of sequelae ranged between € 221 (IBS) and € 22,721 (GBS) per patient per 12 months. Total costs of CE and sequelae extrapolated to Germany 2017 ranged between € 74.25 and € 95.19 million, of which 10–30% were due to sequelae. Conclusion CE is associated with a substantial economic burden in Germany, also due to care-intensive long-lasting sequelae. However, uncertainties remain as to the causal relationship of IBD and IBS after CE.

Comorbidities were assessed by applying the Elixhauser comorbidity index [3,4]. The Chi-square test was used to analyze statistically significant associations with the group assignment. a Three controls per patient, matched by age, gender and state of residence. b One to three controls per patient, matched by age, gender, state of residence and existing comorbidities (in terms of either the Elixhauser comorbidity index [3,4] or the pharmacy-based metric with the ATC algorithm [5], without pain). Chronic conditions were assessed by applying the pharmacy-based metric with the ATC algorithm [5]. The Chi-square test was used to analyze statistically significant associations with the group assignment. a Three controls per patient, matched by age, gender and state of residence. b One to three controls per patient, matched by age, gender, state of residence and existing comorbidities (in terms of either the Elixhauser comorbidity index [3,4] or the pharmacy-based metric with the ATC algorithm [5], without pain).

Goodness-of-fit
McFadden's pseudo R 2 0.16 adjusted R 2 0.13 Results of a two-part regression model: i) probability of non-zero costs via logistic regression, ii) cost estimation via ordinary least squares regression of log-transformed non-zero costs; n=4,036 patients and n=9,350 controls, matched by age, gender, state of residence and comorbidities (in terms of either the Elixhauser comorbidity index [3,4] or the pharmacy-based metric with the ATC algorithm [5], without pain). CE: Campylobacter enteritis, CI: confidence interval.

Goodness-of-fit
McFadden's pseudo R 2 0.10 adjusted R 2 0.06 Results of a two-part regression model: i) probability of non-zero costs via logistic regression, ii) cost estimation via ordinary least squares regression of log-transformed non-zero costs; n=4,036 patients and n=9,350 controls, matched by age, gender, state of residence and comorbidities (in terms of either the Elixhauser comorbidity index [3,4] or the pharmacy-based metric with the ATC algorithm [5], without pain). CE: Campylobacter enteritis, CI: confidence interval.

Model parameter Distribution
Campylobacter enteritis (CE)

Total number of CE cases
69,476 reported cases, age-and gender-specific case numbers [2] no variation

Number of severe CE cases
13,159 hospitalized cases, age-and gender-specific case numbers [6] no variation

Costs of CE per person over 12 months
age-and gender-specific mean costs (Table G); reported CE cases with unknown age and/or gender were valued with the mean costs for moderate CE (Table 3) PERT  [18]; 17-31 % of severe cases remain with permanent disability [9,14] no variation; Uniform(0.17, 0.31)

GBS mortality
-age-and gender-specific mortality calculated as total deaths due to GBS [7] divided by total GBS cases estimated based on country-specific incidence rates [12,19] PERT(deaths divided by lower CI, mean, upper CI of GBS incidence rate) -deaths occur 33 days (0.09 y) after onset of severe GBS [17]; partial costs charged Exponential(1/0.09)

Duration of CD
chronic [20]: remaining age-and gender-specific life expectancy b [18] no variation

Duration of UC
chronic [21]: remaining age-and gender-specific life expectancy b [18] no variation