This was a cross-sectional study. The patients were clinically diagnosed by doctors at IMSS. The data used in this study were obtained from the System of Information of Integral Health Care and the hospital records of the System of Medical Statistics (DataMart) of the IMSS and the databases of Family Medicine Units. At the IMSS, medical attention is classified into three levels: 1st-level facilities perform preventive measures and treat acute and chronic pathologies without complications; 2nd-level facilities address complicated pathologies, chirurgic interventions, and treatments that require hospitalization; and 3rd-level facilities are equipped to treat individuals with complex and complicated diseases. In the IMSS database, records are available for each patient, as well as those who attended medical control in the first, second and third levels of care clinics. Diseases are registered through the codes of the International Classification of Diseases (ICD 10).
The data in the database were compiled by doctors who attend and diagnose patients visiting hospitals affiliated with the IMSS; the data are validated by a team of engineers in computer systems and are analyzed by statisticians affiliated with specialized centers strategically located at regional and state levels.
For this study, patients aged 1 to 24 months hospitalized for severe acute respiratory infection with respiratory distress, febrile syndrome, inadequate oral intake or dehydration from January 2013 to December 2017 in IMSS hospitals were included. Patients included were described using the following ICD-10 codes: B974, which includes RSV; J12 , which includes viral (J120-J129X) or nonspecific (J168, J17, J171, J178) pneumonia, acute bronchitis (J20, J203-J209), and acute bronchiolitis (J21, J210, J211, J218, J219).
The study included patients who, according to records, were first hospitalized at the IMSS for respiratory tract infections.
Patients with immunodeficiency (D81-D84), neurological disorders (G40-G41, G71, G80-G83), airway malformations (Q30-Q34) or incomplete data were excluded.
In-hospital complications were defined as any event or condition detrimental to the patient's health, as caused by an unintentional injury and recorded by the medical and nursing staff during hospitalization. In-hospital complications were grouped and labeled as infectious, respiratory, metabolic and cardiovascular complications.
In addition, codes for respiratory complications such as nonviral pneumonia (J13, J13X, J14, J15, J151, J152, J156, J158, J159, J16, J181, J182), respiratory failure (J96, J960, J969), and nonspecific respiratory disorder (J988, J989) were included in the diagnoses.
An invasive medical procedure was defined as deliberate access to the body through an incision or a percutaneous puncture, where instrumentation is used in addition to the puncture needle or instrumentation through a natural orifice, and performed by trained professionals who use instruments; only invasive procedures related to VRI were included.
Patients were classified according to pathological history, including history of BPD, prematurity and CHD.
A history of prematurity was identified by relevant ICD-10 codes (P070, P072, P073). The patients were subclassified into premature and extremely premature groups according to the history of weeks of gestational age (wGA). Prematurity was defined as 29-36 wGA at birth and extremely premature as <29 wGA at birth. The presence of BPD was identified by code P271, and congenital heart disease with hemodynamic compromise was identified by codes Q20-Q26.
The criteria for severe acute respiratory infection were considered when the patient required endotracheal intubation and was admitted to the PICU. A subanalysis of the patients admitted to the PICU was performed.
Of the included patients, data regarding sex, age, location, month of hospitalization, days of hospital stay and complications during stay were collected, as was the reason for hospital discharge.
The primary outcomes that were included were the length of hospital stay, in-hospital complications, invasive medical procedure and mortality.
Statistical analysis
Quantitative variables are presented as the mean, standard deviation, minimum and maximum. Qualitative variables are presented as proportions and frequencies. Three subanalyses were performed to compare (1) patients with pathological history (prematurity, BPD and CHD), (2) admission diagnoses (pneumonia, acute bronchitis, and acute bronchiolitis), and (3) admission to the PICU.
A nonnormal distribution was observed for quantitative variables (age and length of hospital stay) using the Kolmogorov-Smirnov test; therefore, log-transformation was used for the statistical analysis. Student’s t-test, ANOVA, and χ2 analyses were applied for inferences.
To identify factors related to the length of hospital stay, multiple linear regression was performed; the linear regression model met the assumptions of linearity, normality, homoscedasticity and independence. To identify factors related to in-hospital complications and mortality, a multiple logistic regression model was constructed with "step backward" modeling, and a fitting model was obtained, whereby the noncollinearity of the variables was confirmed.
A value of p <0.05 was considered statistically significant.
STATA v.12.0 was used for the statistical analyses.
The protocol was approved by the National Research Ethics Committee, which belongs to the Mexican Institute of Social Security. This committee is the body in charge of evaluating research projects at the national level.