An administrative data review on hospitalizations due to acute bronchiolitis in children under 2 years of age was performed, from 2000 to 2015, using the anonymized database provided by the Authority for Health Services of the Portuguese Ministry of Health (ACSS) which collects administrative and clinical data in Portuguese hospitalization in mainland public hospitals, including information on diagnoses and procedures performed during a hospital stay, which are coded using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).

Acute bronchiolitis cases with one or more days of LOS were identified using the ICD-9-CM codes 466.11 (RSV bronchiolitis) or 466.19 (unspecified acute bronchiolitis) as main or secondary diagnosis. The percentage of admissions for bronchiolitis was calculated for the total of admissions in children of this age group, before and after excluding those related to routine birth (ICD-9-CM codes V30.x through V39.x as a primary diagnosis).

Variables assessed were age, sex, admission date, discharge date, type of admission, discharge status, other primary and secondary diagnoses, and diagnostic and treatment procedures.

The primary outcome was the rate of hospitalizations due to acute bronchiolitis, to examine the burden that acute bronchiolitis imposes on the national and regional healthcare system and analyze admission trends over time.

Secondary outcome measures were: (1) percentage of patients with risk factors; (2) severity, as assessed by LOS, in-hospital mortality and ventilation requirement; (3) use of diagnostic tests; (4) management strategies; and (5) estimated direct hospitalization charges.

In the trend analyses, hospital admission rates are presented as episode-based of cases per 1000 population, for both sexes combined, aged under <2 years, expressed as an annual rate. The population denominator for calculation of the years in analysis was obtained from INE – Instituto Nacional de Estatística (National Statistics Institute), a public institution responsible for ensuring the production and dissemination of official governmental statistical information.16 Incidence rates were calculated for each year and age category of interest. The denominator values were estimated by the number of live births, using the data of births for the corresponding year. Thus, the number of children aged 0–<3 months was estimated by taking the annual estimate of live births and multiplying by 0.25, as it roughly represents 25% of the children aged <1 year, and the number of children aged 3–11 months by multiplying by 0.75. Time trends regarding the number and rate of hospitalizations were calculated using linear regression models.

Geographical variation was analyzed using Portuguese Authority area boundaries. The place of residence was obtained from the database.

Risk factors were considered if at least one of the following diagnostic codes was present as main or secondary diagnosis: (1) heart disease (congenital or other): ICD-9-CM codes 416.8, 425.4, 746.0x, 745.x, 746.x, 747.0, 747.10, 747.11, 747.21, 747.3x, 747.41, 747.49; (2) chronic lung disease: ICD-9-CM code 770.7; (3) Down's syndrome: ICD-9-CM code 758.0; (4) neuromuscular disorders: ICD-9-CM codes 729.1, 740.x, 741.x, 742.x, 359.X, 335.X; (5) velo-cardio-facial syndrome: ICD-9-CM code 758.32; (6) congenital anomalies of respiratory system: ICD-9-CM code 748.x; (7) other congenital musculoskeletal anomalies; ICD-9-CM code 756.x; (8) immunodeficiencies: ICD-9-CM codes 042, 279.x. We could not assess prematurity data as these specific ICD-9-CM codes are usually not inserted by medical coders, except if the bronchiolitis episode occurs during the first hospitalization.

In-hospital mortality was evaluated based on discharge status and corresponds to the percentage of hospitalization episodes with a primary or secondary diagnosis of acute bronchiolitis which resulted in in-hospital death.

As Intensive Care Unit (ICU) admissions are not separately identified in the database, we used invasive and non-invasive ventilation as surrogate indicators of ICU admissions, since they are not routinely performed in the general paediatric wards. Although there may be incomplete information on diagnostic tests and treatment strategies, and reporting may have varied over time, we were interested in exploring trends in the prescription of diagnostic tests and treatment strategies before and after the publication of national guidelines for acute bronchiolitis.17 The following ICD-9-CM procedure codes were considered: (1) chest X-ray: 87.44, 87.49; (2) microscopic examination of blood: 90.5x; (3) microscopic examination of specimen from trachea, bronchus, pleura, lung, and other thoracic specimen, and of sputum: 90.4x; (4) non-invasive mechanical ventilation: 93.90; (5) continuous invasive mechanical ventilation (through endotracheal tube or tracheostomy): 96.70, 96.71, 96.72; (6) respiratory medication administered by nebulizer: 93.94; (7) oxygen enrichment: 93.96; (8) insertion of (naso-)gastric tube: 96.07; (9) injection or infusion of electrolytes: 99.18; (10) injection of steroids: 99.23; (11) injection or infusion of antibiotics or anti-infectives: 00.14, 99.21, 99.22; and (12) breathing exercise: 93.18.

Charges were calculated from expenditure tables for the Portuguese National Health Service hospital reimbursements, as defined by governmental decree in 2009,18 and were estimated by using a diagnosis-related groups (DRG)-based budget allocation model.

Descriptive statistics were performed using IBM SPSS Statistics 24TM for Windows (Armonk, NY: IBM Corp). Linear regression models with 95% confidence intervals (CI) were performed to assess trends from 2000 to 2015 in the number of hospitalizations, risk factors and diagnostic tests and treatments. Mann–Whitney U tests (mean rank scores) were used to analyze possible differences between LOS and use of mechanical ventilation in patients with or without risk factors. Chi-square tests were performed to assess differences in the use of diagnostic tests and treatments before and after the publication of national guidelines for acute bronchiolitis.

Table 1 shows the hospitalizations due to acute bronchiolitis in children <2 years by year and age group. During the 16 years study period, we recorded 80,491 admissions for acute bronchiolitis in children aged less than 2 years, representing 4.6% of all admissions in this age group and 19.6% after exclusion of admissions related to routine birth. Acute bronchiolitis as a primary diagnosis represented 91.7% of the admissions. The mean yearly admissions for acute bronchiolitis was 5031 cases and ranged from 4308 cases in 2015 to 5820 cases in 2005 (Fig. 1). Children under 1 year accounted for 84.5% of the total of admissions.

Figure 1.

Hospitalizations for acute bronchiolitis in Portuguese mainland public hospitals per year and age group.

(0.13MB).

The mean hospitalization rate was 26.28±2.28 and reached a peak of 30.63 in 2014. In children younger than 1 year the admission rate was 44.64±3.60. Although there was a reduction in hospital admissions during the period of the study, there was an increase in the rate of hospitalizations (B=0.287; r=0.602; p=0.014). Over the time, the total admission rate rose by an average of 1.6% per year, and by 3.75% per year in children younger than 3 months.

North and Lisbon were the regions that presented more cases of hospitalizations for acute bronchiolitis as they are the most populated areas of Portugal. However, the mean admission rate for acute bronchiolitis in Lisbon was 21.84, while it was higher in Centre (27.37) and North (26.38), the northernmost regions, and lower in Alentejo (21.53) and Algarve (19.40), the southernmost regions of continental Portugal.

The baseline characteristics of patients admitted with bronchiolitis are depicted in Table 2. Males were more often hospitalized than females (59.7% vs 40.3%; p<0.001) and unplanned admissions accounted for 79,753 (98.1%) of all hospitalizations. Admissions specifically coded as due to RSV infections represented 39.8% of the total. The presence of any risk factor was reported in 2941 (3.65%) of all admissions. Heart disease (2.3%), Down's syndrome (0.46%) and neuromuscular disorders (0.28%) were the most common underlying conditions.

The mean LOS was 6.1 days and decreased during the period of the study with a maximum of 6.7 days in 2000 to a minimum of 5.5 days in 2014. The median LOS was of 5.0 days [3.0; 7.0]. The children presenting at least one risk factor displayed a higher mean LOS (mean rank 56,022.7 vs 39,736.3; p<0.001).

Fig. 2 shows the percentage of non-invasive and invasive mechanical ventilation use during the period of the study. In recent years, the use of non-invasive ventilation increased, from 0.4% in 2000 to 4% in 2015, while there was no perceivable trend in the use of invasive ventilation. Patients who needed mechanical ventilation had a significantly higher LOS (mean rank 56,465.1 vs 40,170.2; p<0.001).

Figure 2.

Percentage of children admitted for acute bronchiolitis in Portuguese public mainland hospitals who underwent non-invasive (solid line) or invasive (dashed line) mechanical ventilation, per year.

(0.15MB).

We registered 70 cases (0.1%) of in-hospital mortality for infants hospitalized with acute bronchiolitis during the years of the study. The annual rate of in-hospital mortality was stable through the years (0.1%). Patients presenting at least one risk factor had a higher mortality rate compared with those with no risk factor (1.1% vs 0.1%; p<0.001).

Table 3 represents the percentage of diagnostic tests and treatments that patients underwent during the period of the study, respectively before and after the implementation of the national guidelines for acute bronchiolitis in 2012.

The estimated total charges related to bronchiolitis over the period of 2000–2015, including ICU and ward costs, was of 72,420,731.7€, with a mean total charge per year of 4,526,295.7€, and a mean charge per patient of 900.6€. We registered a decrease in the mean total charge per year, 4,298,145.4€ vs. 4,602,345.8€, respectively after and before 2012.