The study is based on the DLCST - a randomized clinical trial - with the overall aim to assess whether CT-screening of heavy tobacco smokers or ex-smokers with a history of at least 20 pack-years can reduce mortality from lung cancer. A total of 4,104 volunteers aged between 50 and 70 years, were included in the DLCST, with randomization at baseline - a specific date during the period 12 May 2004 until 20 June 2006- to either an intervention group (2,052 persons) or a control group (2,052 persons) followed by four annual screening rounds in the period 1 April 2006–31 March 2010. .
A baseline CT scan was performed in all participants of the CT intervention group
Pulmonary nodules were classified according to size, morphology and growth, and the assessment was validated by two radiologists.  At base line (1) nodules smaller than 5 mm and calcified (benign) nodules were tabulated, (2) non-calcified nodules between 5 and 15 mm were rescanned after three months. If the nodule increased in size or was larger than 15 mm the participant was referred for diagnostic procedures. The growth of nodules was assessed by both linear measurement and volumetric analysis.  On an annual basis the CT intervention group was offered a CT scan and completed annual lung function tests together with questionnaires on smoking habits, health status, psychosocial factors, and quality of life issues. The control group had identical procedures as the screening group, but without CT scans.
Persons with a permanent address in Denmark have a unique 10-digit Civil Registration Number (CPR), which has been assigned to each Danish resident since 1968. The number includes information on birthday and sex, and it is used by all authorities for registration purposes. We used the CPR to link the DLCST project data with public registries including The Danish Civil Registration System , The Danish National Patient Register , and with public registries developed by Statistics Denmark such as The Household and Family Statistics (based on The Danish Civil Registration System) as well as The personal income statistics  and The Danish National Prescription Registry of the Danish Medicines Agency. .
Both the intervention group and the control group in the DLCST were followed-up for three years for use of AD or AX, counted from the baseline date of randomization in the DLCST.
Demographic characteristics and co-morbidity of participants
The Danish Civil Registration System contains information on gender, addresses, dates of birth, and death and migration for every person who is or has been a Danish resident at any time between 1968 and the present. Information on civil status and socio-economic position at randomization (study baseline) was obtained from public registries in Statistics Denmark. Civil status was defined as living alone versus all others. Socio-economic position (SES) was defined by level of yearly income and included in the analyses as a categorical variable in three categories (below DKK 250,000, DKK 250–350,000, and above DKK 350,000). Information on co-morbidity was obtained from the NPR with calculation of a slight modification of the Charlson Co-morbidity Index (CCI). [19, 20].
The CCI was re-coded into a categorical variable (0, 1, and above 1 point on the CCI scale). Smoking status at baseline as a categorical variable (current or former smoker) was used as a confounder in supplementary analyses.
Measures of outcome
AD and AX are only available by prescription in Denmark. Using the unique PIN, we obtained complete information about redeemed AD and AX prescriptions by linkage to The Danish National Prescription Registry.  This register covers all pharmacies in Denmark and classifies prescribed pharmaceuticals according to the Anatomical Therapeutic Chemical classification system (ATC) at the level of the generic pharmaceutical. We used prescription of one or more of the following drugs to define the endpoints for the present study: The AD group comprised the overall ATC-group “N06A”, including tricyclic antidepressants (TCA, ATC code N06AA), selective serotonin reuptake inhibitors (SSRI, ATC code N06AB), noradrenalin reuptake inhibitors (NARI, ATC code N06AX) and monoamine oxidase inhibitors (MAO-inhibitors, ATC codes N06AF and N06AG). .
Lithium salts are mostly prescribed for bipolar affective disorders and were not included. Zyban has both an anti-depressive effect (ATC code N 06 AX 12) and is used as a smoking-cessation drug. However, none of the DLCST participants received prescribed Zyban in the follow up period, and therefore Zyban was not included in the outcome. The AX group comprised the overall ATC group “NO5B”: Benzodiazepine derivatives (ATC code N05BA), Diphenylmethane derivatives (ATC code N05BB), Carbamates (ATC code N05BC), Dibenzo-bicyclo-octadiene derivatives (ATC code N05BD), Azaspirodecanedione derivatives (ATC code N05BE), and other anxiolytic medications (ATC code N05BX).
According to guidelines from the Danish health authorities the first-line drug for anxiety disorders in Denmark is SSRI antidepressants. Using prescription of AD or AX as a proxy for medical conditions, without information of the underlying medical condition on the recipes to clearly separate depressive disorders from anxiety disorders, did not allow us to conclude on separate analyses of AD and AX. We have thus presented the results for the analysis of AD or AX as our main results.
First, we compared the screening intervention group with the controls at baseline with respect to demographic characteristics with chi-squared tests, except for continuous variables, which were compared between the groups with Students t-test. Secondly, we analyzed the risk ratio for prescription of AD or AX medication, redeemed at pharmacies at least once during a three year follow-up period, from baseline by proportional hazard regression with adjustment for use of AD and AX in the previous four months before baseline. Thirdly, we analyzed the risk ratio for prescription of AD or AX medication at least once during a three year follow-up period from baseline with exclusion at baseline of users of AD and AX in the previous four months before baseline. In addition, we performed extra analyses by extending the previous use of AD or AX to a two year period, by either analyzing AD and AX separately, or requiring at least two recipes.
The follow-up ended at the date of the prescription of the medication under study, the date of death, date of emigration, date of disappearance, date of a diagnosis of lung cancer, or after 3 years of follow-up for each individual, whichever came first. The Person-time at risk was calculated for each individual. The time during follow-up was counted in days.
Hazard ratios (HR) with 95 % confidence intervals (CIs) were calculated with adjustment for gender, age, civil status, SES, and CCI at baseline, and previous prescription of AD or AX.
In a sub-analysis, and due to the limited statistical power, we used logistic regression to test whether the persons diagnosed with lung cancer in the intervention group more often had used AD or AX compared with persons diagnosed with lung cancer in the control group.
The Danish Data Protection Agency and the Danish Medicines Agency approved the present study (J.nr. 2008-41-2764). The DLCST was approved by the Ethical Committee of Copenhagen County on January 31, 2003 and funded in full by the Danish Ministry of Interior and Health on June 23, 2004. Approval of data management in the trial was obtained from the Danish Data Protection Agency on February 11, 2005. The trial is registered in Clinical Trials.gov Protocol Registration System (identification no. NCT00496977).
All participants gave written informed consent to participation in the Danish Lung Cancer Screening Trial, which was approved by the Ethical Committee of Copenhagen County on January 31, 2003.