INTRODUCTION

In 1956, the World Health Organization (WHO) defined day care as "an organized service for the care of children away from their own homes during some part of the day, when circumstances call for normal care in the home to be supplemented".¹

In Spain, based on population surveys conducted in 2010, 25.01% of women work outside the home and use specialized child-care services.²

Day-care centre attendance is not free from risks to the health of the child,^3,4 and is associated with a considerable increase in the incidence of acute infections during childhood. A higher frequency of acute diseases involves an increased use of health care resources and medication. Thrane et al⁵ showed that the risk of undergoing treatment with antibiotics doubles during the first three months of day care attendance.

The increased incidence of infectious diseases also has an impact on society, as these diseases are transmitted to close contacts in the household.^6,7

All of the above may result in an increased use of health care services, which is already high in the under-five population,⁸ and, considering that there are studies that have shown that past health care use is the best predictor for future health care use,⁹ it would be possible for inadequate use patterns to develop during childhood that would be maintained in adulthood.

Thus, we are facing a reality that has a considerable influence on the everyday health of children and on present and future health care expenditures (as it can lead to a higher risk of iatrogenic injury in children and produce strain on the public health system due to increased use of health care resources).

The aim of this study was to analyse the impact of day-care centre attendance on the demand for medical services and the development of patterns of frequent attendance. We also analysed the morbidity pattern (acute infectious disease and recurrent wheezing) of children that were frequent attenders (FAs).

OBJECTIVES

SAMPLE AND METHODS

We conducted a longitudinal prospective study of two cohorts of children aged 0 to 24 months that differed solely on attending vs not attending day-care centres (exposure factor).

Participants: the study started with the participation of 35 paediatricians and 20 nurses employed in the eight health areas of the Public Health Service of the Principality of Asturias (Servicio Público de Salud del Principado de Asturias [SESPA]), who enrolled 1139 children in the study.

Inclusion criteria: children born between January 1 and September 30, 2010 that receive routine care in the PPC clinics of the SESPA and whose families consented to their participation after being thoroughly informed. Exclusion criteria: presence of respiratory or cardiovascular disease or severe immunodeficiency.

RESULTS

We collected data for 1139 children at 6 months, 1092 at 12 months, and finally 975 at 24 months. The 14.3% of loss to followup was due to employment changes in three paediatricians (7.9%) and residence changes in children included in the sample (7.4%).

Homogeneity of the cohorts under study: both cohorts were homogeneous, except for parental educational attainment and maternal employment (parental college-level education and maternal employment outside the home were more frequent in the cohort that attended day care).

Pneumococcal and rotavirus vaccine coverage was higher in children that attended day-care centres.

Children that attended day care: 43.7% of the children in the study were enrolled in day care at age 2 years (most started attending at age 5 to 6 months). Parental employment was the reason for enrolment in day care in 90.7% of these children.

Number of visits to emergency or PPC services a month in both cohorts (Table 1):

Definition of frequent attenders

We defined FAs as children with a total number of visits in the two years under study that exceeded the mean plus two standard deviations (SDs). Emergency department FAs: the mean number of visits in the first two years of life was 2.69 with a SD of 2.99 (Table 3). Therefore, to be defined as FAs, children had to have made more than 8.67 visits in the first two years of life (5.03% of the total children under study) (Table 4).

Paediatrics primary care FAs: the mean number of visits in the first two years of life was 14.56 with a SD of 10.08 (Table 3). To be defined as FAs, children had to have made more than 34.72 visits in the first two years of life (4.10% of the total children in the study) (Table 4).

The group of children that started attending day care at earlier ages (between 0 and 6 months) had a higher percentage of FAs for both types of setting (Table 5), with a relative risk (RR) of frequent use of emergency services up to six times that of children that did not attend day care, and a RR of frequent use of PPC services that was four times that of children that did not attend day care.

Analysis of morbidity in children that were frequent attenders by number of episodes of disease

When we analysed the mean number of episodes of infectious disease in FAs compared to non-FAs, we found that most FAs did not only seek medical care more often, but also had a higher number of episodes of the most prevalent diseases under study (bronchiolitis, bronchitis, conjunctivitis, gastroenteritis, otitis media, common cold, recurrent wheezing and sinusitis) (Table 6).

When we analysed the morbidity of FAs, we observed that day care attendance was not associated with statistically significant differences in morbidity (Table 7).

Study of medication use by frequent attenders

A higher percentage of FAs compared to non-FAs were treated with oral and inhaled bronchodilators, oral and inhaled corticosteroids, montelukast and antibiotics (P < .05) (Table 8).

DISCUSSION

We conducted a prospective study designed by the research group of the Spanish Association of Primary Care Paediatrics (Asociación Española de Pediatría de Atención Primaria [AEPap])¹⁰ and adapted for the Autonomous Community of Asturias by the Association of Primary Care Paediatrics of Asturias (Asociación Asturiana de Pediatría de Atención Primaria [AAPap]): Infectious disease and health care resource use in children aged less than 2 years that attend day care centres.¹¹ The universe of the study was the PPC clinics of the public health system that routinely serve between 90% and 95% of the paediatric population.¹²

The work was performed on a voluntary basis by 33 paediatricians and 20 nurses, which unquestionably improved data collection but did not allow for random sampling. The study was coordinated by the research group, which collected and reviewed the surveys before submitting them to the statistics team. The data were gathered during routine visits in an attempt to optimise the research capabilities of PPC teams.

The considerable number of paediatricians that participated allowed us to obtain a large sample, but also carried limitations, chief of which was the potential variability in diagnostic criteria. We tried to minimize this bias by predefining the diseases documented in the study.

Another limitation is that the study only took into account visits due to infectious diseases or episodes of recurrent wheezing, a potentially severe disease with a high prevalence in Asturias (45.6% of children experience at least one episode of wheezing in the first 36 months of life).¹²

Each episode that led to a health care visit and a specific diagnosis was documented only once, even if children made more than one visit in the context of a single disease process.

All visits to paediatric primary care and emergency services were documented, even if more than one visit was made for a single episode of disease.

We did not take into account the characteristics of day care centres or the time children spent in them, but in a study of similar characteristics, Lafuente et al¹³ did not find differences in morbidity based on the number of hours spent in day care or the staff-to-child ratio of the different centres.

Although the initial sample included 1139 children, the final sample consisted of 975 (73% of the total children born during the period and that received care in the health facilities under study).

Of the total population under study, 43.7% of children attended day care, a figure that approximated the one published in "Estadística de las enseñanzas no universitarias con 51,8 % de niños escolarizados a los dos años" (Statistics of pre-school, primary and secondary education, with 51.8% of children attending educational facilities by age two years).¹⁴

The mean number of visits to Paediatrics Services (emergency and primary care) per month was higher in children that attended day care and increased significantly when children started attending at an early age. These trends have also been observed in studies conducted in other autonomous communities in Spain.^15,16

The development of the first episodes of infectious disease is displaced to earlier ages, when there is a greater risk of complications requiring medical attention.¹⁷

In our study, the mean number of visits per month to emergency or PPC services by age band (7 to 11 months, 12 to 18 months, and older than 18 months) increased after initial enrolment in day care (P = .000 in all categories). We must take into account that we could not analyse the group of infants aged less than 6 months because we did not have any data for the period preceding day care enrolment.

Although high-frequency use in paediatric age groups is a well-known phenomenon, few studies have been devoted to it.

Defining the number of visits required to categorise a patient as a FA is challenging, and we did not find consistent criteria in the medical literature we reviewed.

The Croatian National Institute of Health¹⁸ considered children to be FAs if the number of visits exceeded the 75^th percentile (more than ten visits a year for a single child). In 2011, Melissa Klein et al set the cut-off point at more than two SD (nine or more visits in one year).¹⁹ Considering that these studies have been conducted in other countries and that the pattern of use of health care services may be influenced by a wide range of factors (personal factors in patients, sociodemographic factors, family or social support factors, and factors specifically related to the health care system [paediatric services versus general medicine at the primary care level, availability and accessibility]), the results of other studies can be used for guidance, but not for comparison.

Studies conducted in Spain have found a prevalence of frequent attendance, defined as the mean plus one SD, of 14.89%, but the age range under study was 0 to 14 years and the duration of followup three months.²⁰ If we apply the same definition to another study conducted in 2012 in the primary care setting with a followup of 12 months, the data show that in children aged less than 1 year, FAs would be those that made more than nine visits a year, and in children aged 1 to 2 years, children that made more than 12 visits. The overall percentage of FAs would be 11.86%.²¹ This study took into account all on-demand appointments and had a significantly smaller sample than our study.

In our study, which analysed visits to emergency services and to PPC clinics separately and had a followup period of two years, FAs in paediatric services corresponded to children that made more than 34.72 visits in the two-year period.

The only factor under study for which we found a significant association with frequent use of services was day care attendance, and this association was stronger in children that started attending day care before age 6 months (RR FA use, 6 for the emergency department and 4.36 for PPC clinics). There are previous studies that support this association.¹⁷

All of the above results in an increased use of health care resources by children that attend day-care centres, not only due to the number of parental work-loss days,²² but also to the increased use of antibiotics,^13,23,24as up to 80% of visits by infants aged up to 12 months resulted in antibiotic treatment.²⁵

We found that the RR of using medication (antibiotics, oral and inhaled bronchodilators, oral and inhaled corticosteroids or montelukast) was higher in children that started attending day care earlier. When we analysed FAs separately (regardless of whether they did or did not attend day care) we found an overall higher use of medication (P < .05).

The term "frequent attender" usually carries negative connotations, so we also set out to determine whether this term could be misleadingly applied to a subset of the children that we serve that have a high-risk immunologic profile (a predisposing factor) and in whom this scheme of care (day care) could act as a trigger, given that day care attendance is associated with an increased incidence of these diseases as well as recurrent wheezing.

Although the total number of FAs in our study was small-which is explained by the fact that the study was not originally designed to analyse this variable-we compared the number and type of infectious diseases in FAs that attend day care versus FAs that do not, and found no statistically significant differences. Therefore, FAs, whether they attend day care or not, are children that on average experience a significantly higher number of episodes of nearly every infectious disease analysed in the study and of recurrent wheezing. This also seems to support the hypothesis of children that are FAs having a specific immunologic profile.

CONCLUSIONS