Cystic fibrosis patient characteristics and healthcare resource utilization in Finland using linked registries

Abstract

Background: Knowledge of prevalence and epidemiology of cystic fibrosis (CF) and healthcare resource use among Finnish people with CF is incomplete.

Methods: We conducted a population-based matched cohort study using retrospective real-world data from linked Finnish national registries. Electronic healthcare data and drug purchases of 102 people with CF were analyzed between January 2015 and December 2019 (follow-up). A 5-fold control population was matched by sex, age, and place of residence. Comorbidities and medication use that occurred at any time during follow-up were assessed; annual rates of hospital service use were adjusted for follow-up.

Results: The prevalence of CF in Finland was 1.85 per 100,000. Median age at diagnosis was 1 year, with 60 % diagnosed at age <2 years and 80 % at age <10 years. Median age at death in people with CF was 31.4 years (n < 5); no controls died. The most common comorbidities included chronic sinusitis (39.2 %), pneumonia (38.2 %), diabetes (20.6 %), and nasal polyps (18.6 %). The most purchased medications were antibiotics (99.0 %) and pancreatic enzymes (84.3 %). The annualized rate of hospital visits was higher in people with CF vs controls (outpatient: mean [SD], 17.4 [14.5] vs 0.9 [3.3]; median, 13.6 vs 0.4, respectively; inpatient: mean [SD], 1.0 [1.66] vs 0.03 [0.14]; median, 0.34 vs 0, respectively).

Conclusions: The prevalence of CF in Finland is remarkably low, likely reflecting unique population characteristics and, in part, delayed diagnosis. Antibiotic use is frequent among Finnish people with CF. Inpatient hospital visits are >30 times higher in people with CF than matched controls.

Keywords: Cystic fibrosis; Finland; Matched cohort study; Retrospective registry study; healthcare resource use.

Fig. 1

Most common comorbidities (A) and purchased medications (B) in the CF cohort (N = 102) and control population (N = 508) over the study period of 2015–2019. CF: cystic fibrosis; ICD: International Classification of Diseases. ^a The proportion test was used to test whether the difference between the CF and control populations was statistically significant. This test was implemented using R prop. test (R Foundation for Statistical Computing; 2018. Available from: https://www.R-project.org/). P values were not adjusted for multiple comparisons. ^b The proportion of people with comorbidities, including pancreatic insufficiency, was based on diagnoses (ICD-10 codes) recorded for the person during the study period. The proportion of people who purchased pancreatic enzymes, which are used to treat pancreatic insufficiency, during the study period is reported in Fig. 1B ^c For the control population, the proportion of people with a comorbidity was calculated with the equation n_yes/(n_yes + n_no), where n_yes is the number of people with that comorbidity and n_no is the number of people without that comorbidity. The proportion of people in the control population with a purchased medication was calculated similarly.

Fig. 2

Outpatient (A) and inpatient (B) hospital visits for the CF cohort (N = 102) and control population (N = 508) over the study period of 2015–2019. CF: cystic fibrosis; IQR: interquartile range; Q1: quartile 1; Q3: quartile 3. ^a Data are presented using box plots, with the horizontal line inside the shaded area indicating the median and the upper and lower borders of the shaded area indicating Q3 and Q1, respectively. Whiskers above and below the shaded area indicate the maximum and minimum values, respectively, of the set without outliers. Top outliers are values larger than Q3 + 1.5 × IQR, and bottom outliers are values smaller than Q1 – 1.5 × IQR. Outliers are not shown in the box plots to preserve anonymity (i.e., a point would represent an individual).