Principle findings
This is a large-scale cohort study that describes the association between area level social deprivation and childhood consultation rates. Between 2013 and 2017, a rise in consultation rate was observed in children who attended general practice services. The trend found in this study is similar to national trends in consultation rates [
27]. Overall consultation rate in children attending general practice services was 3.80 per person-years. Children from most socially deprived areas consulted a GP more frequently for the vast majority of clinical diagnosis compared to children from least socially deprived areas.
For the vast majority of the ICPC-1 chapters, children in SDA group had higher GP consultation rates compared to children in Non-SDA group. However, in the pregnancy/child bearing and family planning ICPC-1 chapter, adolescent females in SDA group were less likely to consult their GPs compared to adolescent females in Non-SDA group. Further analysis revealed that adolescent females from most deprived areas consult their GP less frequently for contraceptives (oral, injection or intrauterine) compared to adolescent females from Non-SDA group. Consequently, confirmed pregnancy, unwanted pregnancy and induced termination of pregnancy were more often observed in adolescent females in SDA group.
Comparison with other studies
With regards to consultation rates in general practice, the findings of our study are largely consistent with the literature. A few studies in the Netherlands show that GPs record between 3 to 7 consultations per child per year [
25,
27,
28]. This is similar to studies performed in other countries, where consultation rates in children range between 2 to 7 per year [
3,
26,
29,
30]. With regards to the effect of socioeconomic factors, our observation that children from most deprived areas consult their GPs 20% more frequently than those from least deprived areas is also largely consistent with the literature. Bruijnzeels et al., [
31] found that the odds of consultation rate was 24% higher in children from low/middle socioeconomic status.
(SES) compared to those from high SES. Elsewhere, Saxena et al., [
3] showed that consultation rates were 18% higher in children from social classes IV-V (lowest social class) than in children from social classes I-II (highest social class). McLeod [
32] observed children aged younger than 6 years and living in the most deprived areas consulted their GP 16% more frequently than those in the same age category living in least deprived areas. In their multivariate analysis, Mukhtar et al., [
7] found an 18% higher consultation rate in the group of patients with an IMD (index of multiple deprivation) score in the 5th quintile (most deprived).
In contrast to these findings, a recent study in England showed that children from most deprived areas accounted for 4.8% fewer GP consultations compared to children from least deprived areas [
8]. The authors argued that this decreased trend was due in part to a shift towards emergency department visits by children from most deprived areas. Given that GPs in deprived area have a 15% increased workload [
33], a shift to unscheduled care (i.e., out-of-hours or emergency care services) is a likely consequence. This was observed in a study were children aged 0–4 years from most deprived areas within urban cities had higher call rates to out-of-hours GPs [
34]. Hence, there is evidence that health seeking behaviors of children and their families are affected by other factors including proximity and urbanization. A 2019 report [
27] concluded that residents in the Netherlands lived on average 1.0 km to the nearest GP. The same report observed that in the city of Rotterdam the proximity between residents and their GPs is on average 0.6 km. Assuming that in our study population, children from different SDA groups had equal proximity to a GP, then it would have been unlikely that our findings would be affected by proximity to GP services.
In order to assure equal access to primary healthcare services, it is important to understand the relationship between area deprivation and health seeking behaviors in the clinical context. Literature on a wide range of reasons for GP encounters amongst pediatric population is scarce [
35,
36]. Of the available literature on area level social deprivation and childhood consultation rates, the majority focus primarily on one clinical diagnosis. Only a handful researched the role of socioeconomic factors and various diagnoses or symptoms according to a validated classification system [
1,
25,
31]. ICPC-1 was first published in 1987 as a tool to assist GPs when they record RFEs, diagnosis/problems or processes. In 1998 the second edition (ICPC-2) was published in part due to a large contribution by the Netherlands, and has been implemented by GPs worldwide [
24]. The RPCD uses a modified extended version of the ICPC-1 managed by the Dutch college of General Practitioners and not the widely used ICPC-2 system. There is considerable overlap between both systems. Furthermore, the ICPC-1 system includes additional 2-digit sub-codes expanding on more diagnoses not otherwise specified in the ICPC-2 system. For instance, symptoms related to contraception is coded as W14.00 in ICPC-2 system. In addition to W14.00 in the ICPC-1 system, the extended version includes W14.01 (pessary occlusive) and W14.02 (contraception injection). Despite the slight differences between ICPC-1 and ICPC-2 systems, our findings have important clinical implications since we included all consultations with a valid ICPC-1 code. To our knowledge this is the first study to report extensively on all ICPC-1 codes and provide accurate consultation rates per body system in a pediatric population taking into account local area deprivation level. We were able to analyse the distribution of multiple diseases across various stages in childhood development. However, contrary to our initial hypothesis, we found that adolescent females from least deprived areas had higher consultation rates within the ICPC-1 chapter pregnancy/family planning. In the Netherlands, prevention is highly emphasized in primary care with Dutch GP’s being the first point of access for contraceptives including oral, injection and placing of intrauterine devices (IUDs) [
37]. This service is covered by the Dutch National health insurance and free for females under the age of 20 years. Furthermore, in the Netherlands, there are very few privately owned family planning clinics since GPs provide this care for free to the general population.
Adolescent pregnancy is regarded as a social problem worldwide not to mention the health risks associated with maternal morbidity and infant mortality [
38]. Information on pregnancy and termination of pregnancy among adolescent females can be used to improve health policy and monitor progress towards reducing these rates which in turn are associated with risk of morbidity and death. Our findings are similar to existing evidence that socio-economic inequality is associated with teenage pregnancies and termination of pregnancy rates [
39‐
41]. Between 1980 to 1990, Smith et al., [
39] found a four to eight times higher rate of consultation for teenage pregnancies in the most deprived postcodes in Scotland. Our study showed a two-fold increase rate in teenage pregnancy in children living in deprived areas compared to those in non-deprived areas. We also observed a three-fold increase in induced termination of pregnancy in female adolescents from deprived areas. On the contrary, Smith [
39] found that teenagers from affluent areas had a higher rate of termination of pregnancies amongst adolescent females. The differences can be in part due to the lack of individual level determinants that could explain why unwanted pregnancies or their termination occur in these groups; such as parental education level, household income, (ineffective) use of contraception, beliefs and attitudes towards unwanted pregnancy [
42].
Strengths and limitations
To our knowledge, this large population-based study to assess clinical and social trends in the use of healthcare by pediatric population in general practice setting in the Netherlands. GPs entered data during or shortly after a consultation which would increase data entry accuracy. However, one limitation is the accuracy of diagnoses in this study. For instance, we were not able to verify the diagnoses recorded by GPs, which could have led to an overestimation and underestimation of certain ICPC-1 codes. Nonetheless, the extraction method used to identify the number of consultations per year per person and the associated ICPC-1 chapter codes was accurate to assess consultation rates which is what we aimed to determine.
Consultation rates is a widely used indicator to determine workload in a general practice setting, however the complexity of consultations was not assessed in this study. For instance we did not explore various tasks performed during a consultation such as rate of prescriptions, referral rates, time spent during consultations (consultation length), investigations or managing comorbidities. Such information would provide a more accurate indication of workload which could be possible in another research using the RPCD, however is beyond the scope of this current study.
Area level social deprivation index or score as a determinant of health outcomes at the individual level is widely used and reported in existing literature. Our study lacked a cut-off point for the SDA index as well as information on individual socioeconomic status for consultation data since children were classified by an area level social deprivation score according to their area code. This may have distorted our findings due to ecological fallacy whereby the relationships we observed in both groups does not necessarily hold for the individual child. Obtaining individual-level data is complex, however, this study provides preliminary insight on the causes of morbidity in children that is driving inequalities at the local area level.