Design
This is a descriptive retrospective observational database-registry study, using data from the Danish National Patient Registry (Patient Registry) and The Danish National Prescription Registry (Prescription Registry). All inhabitants in Denmark are assigned a unique social security number (CPR number) at birth or immigration. This unique social security number is used in all public health registries such as the Patient Registry and the Prescription Registry and for registration of all contacts with the health system, i.e. general practitioners (GP), specialist care out-side hospital, private hospitals as well as examinations and treatment performed in hospital settings. The study was accepted by the Danish Data Protection Agency (ID: no. P-2019-191)
Material
Study population
Inclusion criteria were patients ≥18 of age and below 90 with at least one final ICD-10 Patient registry diagnosis code of DR05/DR059 (cough/cough with no further specification) in the Danish National Patient Registry – and/or – having redeemed ≥2 prescriptions for relevant cough-medication within a 90-day period in the Danish National Prescription Registry in the 10-year period from January 1st 2008 to December 31st 2017. ATC Code: Anatomical Therapeutic Chemical Classification System and International Classification of Diseases (ICD) codes of version 10 was used (ICD-10).
The study consists of three populations from the two national validated databases:
Prescription Registry Population – group 1: Included in the prescription registry population are patients who have redeemed ≥2 prescription medications within 90 days to treat cough (Prescription registry ATC code R05) and including ATC codes for Codeine (R05DA04), Noscapine (R05DA07), Dextromethorphan (R05DA09), Pectyl (R05FA02), Opium drops (A07DA02) and all mucolytics (R05CB). Furthermore, for the identification of co-morbidities possibly relating to cough, a search for upper and lower respiratory drugs (Prescription registry ATC codes R03 and R06) and reflux (ATC Patient registry: A02) were identified. The Prescription registry population represents primary care praxis since most cough related medications will be prescribed by general practitioners.
Patient Registry Population - group 2: Included in the population based on diagnosis from the patient registry are patients who have been diagnosed with the cough-specific diagnosis (DR05/059) in the National Danish Patient Registry.
The identification of co-morbidities possibly relating to cough were done by searching the diagnosis registry for chronic lower respiratory disease coded with DJ40-46, including COPD (DJ44) and asthma (DJ45). Acute lower respiratory diseases (ICD-10 Patient registry of DJ20-22), upper respiratory illnesses (ICD-10: DJ30-39) and lastly, GERD was based on ICD-10 code DK21, and psychogenic cough on ICD-10 code DF453. Non-respiratory illness which might lead to cough, such as heart failure (DI24-25), different types of arthritis (DM051-52, and DM060-69), pulmonary embolism (DI26) and diabetes (DE10-11) were examined as well. The Patient registry population represents secondary health care from hospital/specialist setting.
Combined Prescription- and Patient Registry Population – group 3: A mergedpopulation with patients present in both the Prescription Registry and Patient Registry databases using the above inclusion criteria which represents the group of patients who have been in contact with both primary and secondary health care.
Exclusion criteria was the use of ACE-inhibitors (C09AA) within 12 weeks of diagnosis date, antibiotics (J01) within 6 weeks of diagnosis date and any malignant respiratory disease in the larynx (ICD-10 C32), tracheus (C33) or bronchus/lung (C34).
Registries used.
The Danish National Patient Registry records all ICD-10 diagnoses given within the Danish healthcare system which enables tracking of all patient contacts (12). There is a strong economic incentive for the hospital departments/specialists to use the system, since the Danish National Patient Registry uses the ICD-10 and testing codes for reimbursement (12).
The Danish National Prescription Registry records all specific prescriptions’ ATC codes redeemed from Danish pharmacies also using the unique social security number. Over the counter medications are not recorded in the prescription registry and are therefore not possible to include in this study.
The Prescription registry population (group 1) is allocated from the first event of ≥2 prescriptions redeemed within 90 days in the 10-year study period. The 90 days were selected, due to the content of the package of cough medication and the devices in respiratory medicine. Tablets towards cough are most often prescribed with 40 tablets per package, and 2 prescriptions are needed within less than 90 days. In case of 3 tablets per day, a package of 100 tablets would last 33 days and in case of 2 tablets daily the amount will last 50 days. Mixture of cough medication would be prescribed in 200 mL per bottle, with the lowest doses of 5 mL, lasting 40 days. Respiratory devices containing 30 doses for treatment once daily and 60 doses for treatment twice daily, i.e. one device per month, some packages might include 3 devices, which cover treatment for 90 days, but not 91 days, therefore number 2 prescriptions would be needed within 90 days, for continuation of treatment and this cut-off was selected in the current survey. The Patient registry population (group 2) is based by first occurrence of a final ICD-10 codes from a contact in the patient registry for cough-diagnoses (DR05 and DR059) and therefore patients only occur once in either population.
Cough specification: Chronic cough is defined as cough >8 weeks (11). In this study the definition used for chronic cough is ≥2 redeemed prescriptions within 90 days or having a contact with the healthcare system where final diagnosis of the contact being ICD-10 code DR05/DR059. A final diagnosis code of cough from a health care contact can be used as a proxy for chronic cough since the patient will be referred by a general practitioner after a longer period of coughing before the hospital contact. The other reason is the economic incentive from the hospital/specialist to use a ‘real’ ICD-10 diagnosis code for reimbursement and only use cough if no other diagnosis is present. To validate this study’s chosen proxy on chronic cough an analysis of the Patient registry sub-population with a contact of ≥8 weeks and then final diagnosis code DR05/DR059 was also performed.
The included populations were identified as having either Possible Unexplained Chronic Cough (UCC) or Possible Refractory Chronic Cough (RCC). Possible UCC being defined as not having a cough-relevant co-diagnosis within 12 months before and after the point at which the diagnosis date was registered and possible RCC being defined as having a cough-relevant co-diagnosis within 12 months on either side of cough diagnosis-date. For the Prescription registry population co-morbidities were identified through other prescriptions if they did not also have a contact in the Patient registry.
Outcomes of the study:
The outcomes of the study were pre-defined and were as follows:
Primary Outcome: A descriptive table showing baseline characteristics and prevalence of individuals with possible UCC– including demographic composition and characteristics as well as relevant examinations in relation to diagnosis.
Secondary Outcomes: A descriptive table showing baseline characteristics of individuals with possible RCC – including demographic composition and characteristics, cough-related comorbidities, as well as relevant concomitant medications and relevant examinations in relation to diagnosis.
Demographic data such as age, gender, and living region in Denmark (Capital city area, urban area, rural area) is registered in the Patient registry, whereas hospital-based prescription and hospital-based treatments such as biological drugs is not accessible through the Prescription Registry and is therefore not reported if the patients are only registered in the Prescription Registry. The cough-relevant medication redeemed at the pharmacies for inclusion in the prescription-based population is reported (see below). (1) Respiratory medication R03, antitussive and cold medication (R05), antihistamine (R06), and lastly, opium-drops (A07/DAO2
Examinations and tests: For the Patient registry population a search for relevant examinations related to cough, such as lung function testing, asthma provocation, chest x-ray, CT-scan of thorax, HRCT of the lungs, gastroscopy, laryngoscopy, and bronchoscopy was performed (360 days on each side of diagnosis-date) to describe the relevant examinations in connection with cough diagnosis-date. It is not possible to perform a search for relevant examinations for the Prescription registry population; therefore this is not reported. There was no access to patient records and blood tests.
Statistics
Data analysis was performed using SAS for Windows (SAS, Cary, NY, US) version 77.1. Categorical variables are described as absolute numbers as well as percentage where possible. All data generated or analysed during this study are included in this published article. The analysis for demographic data included all inhabitants from Denmark, based on prescription lists and ICD-10 Patient registry’s, co-diagnosis, co-medications, and relevant testing as well as examinations were performed for the population of the primary (UCC) and secondary outcome (RCC). A merge of the Prescription registry population and the Patient registry population within both UCC and RCC was performed.