Methods
The study, which is a retrospective cohort type, started in May 2020
under the direction of Marmara University Health Sciences Institute
Medical Pharmacology Department and was carried out between May 2020 and
October 2021. Patients registered at Istanbul Uskudar Zeynep Kamil
Family Health Center (FHC), Sultanbeyli Jandarma Ustegmen Rahim Celik
FHC, Beyoglu 6th FHC and Eyup Islambey FHC were included in the study.
Of these patients, those who were diagnosed with COVID-19 between March
11, 2020 - November 30, 2020, constituted the population of the study.
Inclusion criteria were to be a patient enrolled in the above-mentioned
FHCs and to have a positive COVID-19 PCR test between 11 March 2020 and
30 November 2020 or positive COVID-19 thoracic CT findings despite a
negative test. The exclusion criterion from the study was the inability
to access the medical information of the included patients viae-nabiz (an application that Turkish citizens and health
professionals can access health data collected from health institutions
via the internet and mobile devices).
The records of the patients included in the study were scanned
retrospectively. Data scanning was performed via family medicine
information systems and e-nabiz . The data of the patients were
collected by the researcher working in the relevant FHC and
participating in the study. The information obtained from the patient
files were as follows: patient’s age, gender, marital status, education
level, employment status, occupation, smoking habit, date of first
diagnosis, first application complaint, COVID-19 PCR test result,
presence of COVID-19 findings in thoracic CT, received COVID-19
pharmacological treatment, presence of comorbidity and medications used
in the previous 6 months. The medications utilized in the previous 6
months were recorded according to the Anatomic Therapeutic Chemical
(ATC) 5 classification. It was then further grouped according to ATC 3.
These data were compared with the three main endpoints of the study,
namely, the need for hospitalization, the need for intensive care unit
(ICU), and mortality in comparative analyses.
SPSS 21.0 was used for statistical analysis. Frequency analysis was
performed by specifying numbers and percentages for categorical
variables. Normal distribution was tested using Kolmogorov–Smirnov
test. Mean and standard deviation were used for continuous variables,
and median and range of values were used for non-parametric variables,
as measures of central tendency and dispersion, respectively. Logistic
regression was used to compare the independent variables with the three
dependent variables. The direction of the comparisons was confirmed by
correlation analysis and Pearson chi-square test. All logistic
regression models were adjusted for age and comorbid conditions.
Demographic data were also compared between outpatients,
hospitalization, ICU admission and mortality using Pearson-chi square
test, when needed Fisher’s exact test. For parameters with nonnormal
distribution, ranks were compared using Kruskal–Wallis test. For
normally distributed parameters, means were compared using one-way
analysis of variance (ANOVA) test. In the presence of significant
variables Tukey’s post hoc test was performed after ANOVA test, Dunn
test was performed after Kruskal-Wallis test and Pearson chi-square or
Fisher’s exact test in pairs were performed after Pearson chi-square or
Fisher’s exact test.
To obtain study data from family medicine information systems ande-nabiz , permission was obtained from the Turkish Republic
Ministry of Health General Directorate of Health Services Scientific
Research Platform on 29.04.2020. Before starting to collect the study
data, an application was made for the approval of the Marmara University
Faculty of Medicine Clinical Research Ethics Committee and the ethics
committee approval was obtained on 08.05.2020 with the protocol code
09.2020.552. The study was carried out in accordance with the principles
in the Declaration of Helsinki.