Cancer is the primary cause of morbidity and mortality throughout the world. The World Health Organization (WHO) estimates that the overall burden of cancer is increasing, and predicts that in 20 years1 it will account for over 50% of all morbidity and mortality. Lung cancer (LC) is the third most common cancer in the world, the most common if both sexes are taken together, and the malignancy with the highest mortality, with 1824701 new cases being diagnosed every year worldwide, leading to 1589925 deaths every year.2

In Spain, it is estimated that 27859 new cases were diagnosed in 2014, 22455 in men and 5404 in women. In men, it was the third most common cancer after prostate and colorectal cancers, and in women, the fourth after breast, colorectal and uterine cancers. Age-adjusted incidence rates standardized to the European population were 75.8 per 100000 and 16.6 per 100000 for men and women, respectively.3

Diagnosis has improved spectacularly in the last 30 years with the use of methods such as computed tomography (CT) and positron emission tomography (PET)-CT,4,5 but diagnosis is often made when the disease is in advanced stages, and 5-year survival rates remain low. LC has traditionally been a man's disease, but disease rates among women have risen sharply since they started smoking, an increase that has been attributed, among other things, to hormonal factors and an increased sensitivity to tobacco smoke.6–8

According to the literature, mean age at diagnosis is between 65 and 75 years.2,6,7,9–11 However, technological advances have extended the survival of LC patients, causing a direct increase in the associated comorbidity burden.6,11

Several studies have indicated that 50%–80% of the healthcare costs of these patients is associated with hospital admissions, while only 20% goes on diagnostic procedures, medical interventions, and palliative care.12 However, very few studies indicate the specific costs of hospitalizations for LC, or the procedures most frequently performed in these patients.

The aim of this study was to analyze changes in the incidence, clinical characteristics, diagnostic procedures, comorbidities, length of stay, and costs in patients hospitalized for LC in Spain between 2001 and 2011.

We performed a retrospective study using the National Epidemiological Surveillance System for hospital data (minimum basic data set [MBDS]) managed by the Ministry for Health, Social Affairs and Equality. This system uses the clinical codes of the Spanish version of the 9th International Classification of Diseases (Clinical Modification International Classification of Diseases [ICD-9-CM])13 for data management and coding. The following data are collected annually: patient details (age, sex, and autonomous community of residence), episode (date of admission, type of admission, type of hospital, hospital group, date of discharge, type of discharge, readmission, and cost), and various clinical variables (primary diagnosis, 14 secondary diagnoses, and procedures performed).

All patients hospitalized in Spain between 2001 and 2011 with a primary diagnosis of LC were recruited (codes ICD-9: 162, 162.2, 162.3, 162.4, 162.5, 162.8, 162.9 and 231.2).

In this study, we detected 298435 hospitalizations for LC in Spain between 2001 and 2011, with an increase in the overall rate of admissions during that period. Although few studies provide numerical data on hospitalizations in LC patients, it is worth mentioning the work of García et al.,15 who reported on 335186 hospitalizations for cancer in Spain in 2008. In our study, 28543 patients were hospitalized for LC during that year.

We also found a marked difference between men and women, with a reduction in the rate of hospitalization among men, and an increase among women. These data coincide with the general trend described in other studies, such as that of Bello et al.,1 conducted between 1995 and 2006. Egleston et al.8 reported a LC risk of 1 in 13 among men and 1 in 16 among women, and described a geographical distribution of the risk which was higher in Europe than in Asia and Australia. The incidence of LC among women has increased in the last 20 years, due, among other reasons, to the increase in smoking among women. Chang et al.2 noted this trend when they revealed a fall in the men-to-women incidence ratio from 2.16 in 1996 to 1.91 in 2008. Similar data were published by Galceran et al.3 who described a lower rate of hospitalization among men, probably due to a reduction in the incidence of LC in this subject subgroup, which fell by 6.1% a year in the period 1993–2014. In contrast, in Spain an increase of 48% was observed among women over the same period.3

The age group with the greatest incidence in our case was the 70–79-year-old group; in this group, no difference was observed between the sexes. Our data are in line with the mean age (72.1 years) of LC diagnosis in Spain reported in other studies,10,11,16 although women were younger at diagnosis than men.17 In other countries, such as Taiwan and Croatia, a younger mean age of 67.8 years has been reported, with an increased incidence of LC among patients aged between 70 and 79 years.2,18 However, we must remember that our study refers to the mean age of patients at the time of hospitalization, not only age at diagnosis, but also in patients who were admitted for other reasons apart from their diagnosis of LC and/or LC complications.

In the 10-year study period, we observed a reduction in the mean length of hospital stay, down to 10.07 days in 2011, and an increase in the mean cost of admission, up to €5562.54. It must be taken into account that the length of stay includes days that the patient remained in hospital for diagnostic procedures and/or for LC treatment, the management of complications, and performance of procedures. Although few studies have been published regarding hospitalizations in these patients, García et al.15 reported a mean length of stay of 11.2–11.4 days in patients admitted for LC attributable to occupational exposure in 2008, with a cost per admission of €5728.91–€5777.07, although they did not specify if the hospitalizations were scheduled for diagnostic or other procedures. Skaug et al.12 published a study of hospital stays in LC patients between 1990 and 1996, with follow-up until 2003. They classified hospital stay into three groups: diagnostic hospitalization, terminal hospitalization (patient admitted for palliative care including end-of-life care), and other hospital stays. They reported a mean stay of 7.8 days, with a total 35 days hospitalization per patient throughout the study period. Mean length of stay figures according to their classification were as follows: 18.5 days for diagnostic hospitalization, 21 days for terminal care, 27.5 days for palliative care, and 57 days for hospitalizations for other reasons. With regard to costs, Demeter et al.18 described the distribution of the annual costs of patients with a diagnosis of LC: 30% was spent on home care, 24% for long-term institutional care (nursing homes, etc.) and 46% for hospital management, including diagnostic procedures and/or treatment.

Most patients in our study had a low comorbidity burden (Charlson 1–2), although the percentage of patients with Charlson>2 increased during the study period. Other studies underline the importance of comorbidities as a prognostic factor in the diagnosis and treatment of LC, and report a comorbidity burden that increases with age at the time of the finding. However, we could not establish any specific relationship between comorbidities and how they were divided numerically.

All admissions registered in this study occurred in hospitals belonging to the National Health System, until private hospitals were incorporated in 2005, which by 2011 accounted for 9% of all hospitalizations. We have not found any other similar studies with which to compare our data.

The procedure most frequently performed during the hospital stay was chest CT, at a rate of 74.8% in 2011. These data coincide with the percentages registered in the published series. This examination and PET-CT are the procedures which have done most to improve diagnosis, staging and follow-up in LC.5,8,19

In the multivariate analysis of trends and factors associated with the incidence of hospitalization for LC during the study period, we observed a statistically significant reduction in incidence in all years except the 2009–2011 period. Factors associated with the incidence of admissions were age for patients aged 70–79 years, compared to those aged<40 years, and female compared to male sex. A similar reduction in incidence over the last 10 years has been highlighted by other authors, interestingly in the same age group and with same difference between sexes.2,7,10

This study has several limitations. Firstly, the information collected in the database has certain methodological problems (including lost data and transcription errors): although the database contains a wealth of information, access to data that would be of great clinical interest is limited, e.g. smoking habit and lung tumor histology (several histologies in the same patient cannot be differentiated), and it is impossible to establish in which hospital admission the diagnosis was made. Moreover, the relative lack of literature and the absence of similar studies to ours have restricted our ability to set up comparisons and parallels.

In conclusion, between 2000 and 2010, the hospitalization rate fell among men and rose among women, the comorbidity burden and cost of LC increased, and mean hospital stay became shorter. Factors associated with the incidence of admissions for LC are age, sex, and year of study: patients were predominantly in the 70–79-year age group, and with regard to sex, women predominated over men.

The authors declare that they have no conflict of interests directly or indirectly related with the contents of this manuscript.