Association between sociodemographic status and antiepileptic drug prescriptions in children with epilepsy

Authors


Address correspondence to Peter Mattsson, Department of Neuroscience, Neurology, Uppsala University Hospital, SE-751 85 Uppsala, Sweden. E-mail: peter.mattsson@neuro.uu.se

Summary

Purpose:  We investigated whether in Sweden sociodemographic differences are associated with access to expert health care and antiepileptic drug (AED) prescriptions in children with epilepsy.

Methods:  Data on epilepsy, prescription of AEDs, and sociodemographic variables were obtained from several national administrative registers. We linked individual data to examine whether access by pediatric epilepsy patients to neuropediatricians and the prescription of individual AEDs differed according to gender, age, parental education, place of residence, parental region of birth, and household income. We also assessed whether AEDs are prescribed differently to patients with epilepsy by neuropediatricians as compared to other physicians.

Key Findings:  Of 1,788,382 children aged 1–17 years in 2006, living in the country by the end of 2006, 9,935 had a diagnosis of epilepsy (0.56%). Patients with epilepsy on AED treatment (n = 3,631) comprised 0.24% of the total Swedish population aged 1–17 years. Out of 3631 patients with epilepsy on AED treatment, 2301 (63.4%) received prescriptions from a neuropediatrician. Children with epilepsy aged 1–5 years old—as opposed to older children and adolescents—and children with epilepsy residing in large cities—as opposed to children living in smaller cities and rural areas—were more likely to be treated by a neuropediatrician. Children living in large cities received oxcarbazepine to a greater extent than children living in rural areas. Levetiracetam was prescribed more extensively to children whose parents had higher incomes. Of the five most frequently used AEDs, three (lamotrigine, oxcarbazepine, and levetiracetam) were prescribed to a larger extent by a neuropediatrician rather than by other specialists, and one AED (carbamazepine) was prescribed to a lesser extent.

Significance:  The results of this nationwide cross-sectional study of children with epilepsy are important because they show that universal coverage for medical care does not eliminate inequalities of access to health care services among children and adolescents. No data are available that can guide us as to whether the density of child neurologists is of importance to access to expert health care, but this seems likely. Prescription patterns of AEDs differ between child neurologists and other specialists.

Health equity can be defined as the absence of systematic disparities in health between more and less advantaged social groups or absence of health inequalities that are unjust or unfair (Braveman & Gruskin, 2003). Inequalities may include differences in the presence of disease, health outcomes, or access to health care across racial, ethnic, sexual orientation, and socioeconomic groups. Equity is an ethical principle, consonant with and closely related to human rights principles.

Access to medical care in general differs among European countries and among sociodemographic groups, where the rich are more likely to see a specialist than the poor in most countries (van Doorslaer et al., 2006). Provision of epilepsy care is known to vary between European countries (Malmgren et al., 2003), but less is known about inequalities in epilepsy care between sociodemographic groups within a country.

In September 2011, the European Declaration on Epilepsy was passed in the European Parliament (2011). The declaration called on the Commission and Council to take initiatives to encourage Member States to ensure equal quality of life, for example, in education, employment, transport, and public health care, for people with epilepsy. According to the Swedish Health Care Act, patients should be provided with the health care they need, regardless of sociodemographic belonging.

Identifying existing disparities is the first step in enabling the health care system to take actions to fight unnecessary inequalities. In a previous study by us (Mattsson et al., 2010) there was evidence that adult patients with epilepsy in Sweden were provided epilepsy treatment according to their sociodemographic belonging, and not only according to their needs. This prompted us to study whether there are sociodemographic differences also in the provision of specialized health care for children with epilepsy.

We used a nationwide sample and linkage to analyze whether access to neuropediatricians and prescription of antiepileptic drugs (AEDs) are associated with sociodemographic variables in children with AED-treated epilepsy. A further objective was to assess whether AED prescription patterns in epilepsy differ between neuropediatricians and other medical specialists.

Methods

Information on gender, age, country of birth and residence was obtained from the Total Population Register.

Study population

The number of children aged 0–17 years in 2006 who were living in the country by the end of 2006 was 1,933,918. We had information about how families were composed in 2005, that is, it was possible to link different members of a household, for example, children to their parents. We could not link children born in 2006 (aged 0 years) to a family, so they were excluded. This left 1,788,382 (92.5%) children for further analyses.

We identified epilepsy patients in the National Patient Register (NPR), which provides reliable International Classification of Diseases, Ninth and Tenth Revisions (ICD-9 and ICD-10) codes for diagnosis for all public, in-patient care since 1987, and ambulatory care in hospitals since 2001. We obtained information about AED treatment from The Swedish Prescribed Drug Register (SPDR), established in 2005, which contains data on patients’ unique identifiers, all the prescribed dispensed drug, date of prescribing and dispensing, and the prescriber’s specialty. Drugs are classified according to the Anatomical Therapeutic Chemical (ATC) classification system.

The number of children aged 1–17 years in 2006 and living in the country by the end of 2006 with a diagnosis of epilepsy (ICD-10 codes G40 and G41 during 1997–2005 and ICD-9 codes 333C [progressive myoclonic epilepsy] and 345 during 1989–1996) was 9,935 (0.56% of 1,788,382). This was the study population.

Of 9,935, there were 4,373 (44.0%) children on continuous AED treatment, defined as dispensings of one or more AEDs both in January–April and September–December 2006. In Sweden the maximum quantity of drugs dispensed is for 90 days of supply. Therefore, the study period of 4 months should include filling of most prescriptions. Of 4,373 children receiving continuous AED treatment, 3,631 (83.0%) had a complete data set. This was the effective sample. There were 742 (17.0% of 4,373) dropouts, that is, they had missing data either on the prescriber’s specialty (n = 715) or on parental education (n = 27). There were 832 children on AED treatment without a diagnosis of epilepsy.

Definitions of variables

Special needs for rehabilitation services

The following diagnoses were considered markers of the presence of rehabilitation needs (ICD-9 codes followed by ICD-10 codes): mental retardation (377–319, F70–F79); disorders of psychological and motor development (315, F80–F83 and F85–F89); autism (299, F84); cerebral palsy and hemiplegia (333H and 342–344, G80–G81); congenital malformations of the nervous system (740–742, Q00–Q07); and chromosomal abnormalities (758–759, Q90–Q99). Children with any of these diagnoses in 1989–2005 were identified in the NPR using ICD codes and were considered to have special needs for rehabilitation services. In Sweden, the specialty of neuropediatrics includes responsibility of pediatric neurology and of rehabilitation medicine. Children with rehabilitation needs are therefore more likely to have their prescriptions from neuropediatricians.

Comorbid disease

Children with comorbidities—other than those directly related to rehabilitation needs—recorded in 2004–2005 and relevant to this study were also identified in the NPR. We argued that children with a comorbid disease might have a greater chance of having their basic care from neuropediatricians. The following diagnoses were considered markers of the presence of relevant comorbidities (ICD-9 codes followed by ICD-10 codes): diseases of the nervous system (320–359, not including 345, 333C, 333H or 342–344, G00–G99, not including G40–G41 or G80–G81); stroke (430–438, I60–I69); metabolic disorders (270–279, E70–E79); intracranial injuries (850–854, S06); benign neoplasm of meninges, brain, and other parts of central nervous system (225, D32–D33); malignant neoplasm of meninges, brain, and other parts of central nervous system (191–192, C69–C72); and hyperkinetic disorders (314, F90). Children with any of these diagnoses in 2004–2005 in the NPR were considered to have comorbidities.

Access to neuropediatricians and prescriptions of AEDs

Any dispensing by a neuropediatrician defined the AED user as having access to a neuropediatrician. In 2005 there were 96 clinically active neuropediatricians in Sweden (2007). One or more dispensings of any AED with an ATC code = N03 during year 2006 defined the patient as being prescribed that AED.

Sociodemographic variables

Information on individual educational levels in 2005 was obtained from the Swedish National Education Register with information on the highest formal education attained by each parent.

Educational level was represented at three levels: elementary (9 years or less of schooling), secondary (10–12 years), and university level (>12 years). Place of residence was divided into the following: (1) large cities (Stockholm, Gothenburg, and Malmo), (2) middle-sized cities (places with >90,000 residents within 30 km of the municipal center), and (3) rural municipalities (all other areas). Region of birth was divided into Sweden and other countries. Age was divided into three classes, where age-related differences in health are expected to occur.

We had access to the individuals’ total family disposable income for 2005 from the Total Enumeration Income Survey. Disposable income includes earnings and wage-related benefits and cash benefits, less annual income tax, accounting for household size and structure. The households of children with an epilepsy diagnosis and continuous AED use were used as the basis for calculating ranges for income quintiles. All children in the study were then classified in quintiles according to their families’ disposable income.

Analyses and statistics

Differences in distributions in the effective sample were analyzed by maximum likelihood methods:

  • 1 Logistic regression was used to calculate crude and adjusted odds ratios with 95% confidence intervals (CIs) as estimates of whether access to prescriptions by neuropediatrician differed between sociodemographic subgroups of epilepsy patients.
  • 2 Logistic regression was used to calculate crude and adjusted odds ratios with 95% CIs as estimates of whether access to prescriptions of the five most prescribed AEDs differed between neuropediatricians and other specialists. Prescription patterns of other AEDs were not analyzed.
  • 3 We then compared the prescription patterns of the five most prescribed AEDs within each sociodemographic variable using crude and adjusted odds ratios with 95% CIs. Prescription patterns of other AEDs were not analyzed.

The following sociodemographic independent variables were of interest: gender, parental education, age, place of residence, parental region of birth, and household income. When one of them was analyzed the others were included as confounding variables together with comorbidity and rehabilitation needs (the latter as dichotomous yes/no variables). To control for unobserved heterogeneity related to between-region variation present in the data, we also included fixed effects for counties in our regressions.

Findings are considered important and are provided in the Results section when the 95% CI of the estimates did not include 1 and the odds ratios were either <0.8 or >1.2. We put focus on the adjusted estimates because in interpreting the effects from each of the studied sociodemographic variables we wanted to exclude possible confounding effects of the other variables.

The statistical program Statistical Analysis System (SAS), version 9.1, (SAS Institute Inc., Cary, NC, U.S.A.) was used in all analyses. The study was approved by the regional research ethics committee.

Results

Patients with epilepsy on continuous AED treatment comprised 0.24% of the total Swedish population aged 1–17 years; of these 63.4% received care by a neuropediatrician.

Dropouts differed from children in the effective sample by to a larger extent living in rural areas, lower parental incomes, and less comorbid disease and rehabilitation needs (Table 1). AED users without a diagnosis of epilepsy differed from children in the effective sample in the following ways; the former were younger, lived to a larger extent in rural areas, their parents were born in Sweden to a larger extent, their parental incomes were lower, they had less access to neuropediatricians, and they had less comorbid disease and rehabilitation needs (Table 1).

Table 1.   Distribution of characteristics in children with epilepsy on AED treatment, dropouts, and in children on AED treatment but without any epilepsy diagnosis
 Epilepsy and AED treatment (n = 3,631)% of 3,631Dropouts (n = 742)% of 742AED treatment but no diagnosis of epilepsy (n = 832)% of 832Statistical comparison
  1. Dropouts: no information available on specialty of prescriber; AED, antiepileptic drug; na, not applicable because no data were available; ns, no statistically significant difference between the distribution of the three groups (effective sample, dropouts, and children on continuous AED treatment but without epilepsy diagnosis).

  2. ap < 0.01 for comparison between all three groups.

  3. bp < 0.01 for comparison between effective sample and dropouts.

  4. cp < 0.01 for comparison between effective sample and children on continuous AED treatment but without epilepsy diagnosis.

Gender       
 Boy1,91652.840154.046756.1ns
 Girl1,71547.234146.036543.9
Parental education       
 Missing data00.0273.600.0ns
 Low1,30235.927637.231137.4
 Middle1,40938.828037.731237.5
 High92025.315921.420925.1
Age (years)       
 1–544512.3739.817821.4 a,c
 6–121,59443.933945.734040.9
 13–171,59243.833044.531437.7
Place of residence       
 Rural/small city1,01427.924132.529635.6 a,b,c
 Middle-sized city1,26834.930140.631337.6
 Large city1,34937.220027.022326.8
Parental region of birth       
 Other country (both)50814.09512.8789.4 a,c
 Sweden (at least one)3,12386.064787.275490.6
Parental income (quintiles)       
 1 (low)71619.716121.722426.9 a,b,c
 271019.616522.215919.1
 370219.317022.919723.7
 474520.513117.713516.2
 5 (high)75820.911515.511714.1
Comorbidity       
 No2,42866.956776.467180.6 a,b,c
 Yes1,20333.117523.616119.4
Access to neuropediatrician       
 No1,33036.6na 49259.1 c
 Yes2,30163.4na 34040.9
Needs for habilitation       
 No1,71147.151969.958270.0 a,b,c
 Yes1,92052.922330.125030.0

Sociodemographic differences in access to prescriptions by specialists

Children aged 1–5 years, children living in large cities, and children with rehabilitation needs were more likely to be treated by a neuropediatrician (Table 2).

Table 2.   Percentages and odds ratios (ORs) for being treated by a neuropediatrician versus other physicians in different sociodemographic subgroups among continuous AED users with a diagnosis of epilepsy (N = 3,631)
Variable%aCrude OR (95% CI)Adjusted ORb (95% CI)
  1. CI, confidence interval.

  2. aPercentage of patients treated by a neurologist.

  3. bSeparate odds ratios are adjusted for all the other variables in the model and county.

Gender   
 Boy62.91.00 (ref)1.00 (ref)
 Girl63.81.04 (0.91–1.19)1.10 (0.94–1.29)
Parental education   
 Low60.91.001.00
 Middle63.51.12 (0.96–1.31)1.13 (0.94–1.36)
 High66.61.28 (1.07–1.53)1.04 (0.84–1.30)
Age (years)   
 1–567.41.001.00
 6–1263.00.82 (0.66–1.03)0.72 (0.56–0.94)
 13–1762.60.81 (0.65–1.01)0.71 (0.55–0.93)
Place of residence   
 Rural/small city49.31.001.00
 Middle-sized city57.61.40 (1.19–1.65)1.08 (0.87–1.34)
 Large city79.33.94 (3.29–4.72)1.55 (1.15–2.07)
Parental region of birth   
 Sweden (at least one)62.01.001.00
 Other country (both)71.91.56 (1.27–1.92)1.05 (0.82–1.34)
Parental income   
 1 (low)62.71.001.00
 258.90.85 (0.69–1.05)0.99 (0.77–1.28)
 362.30.98 (0.79–1.22)1.08 (0.84–1.39)
 463.41.03 (0.83–1.27)1.12 (0.87–1.45)
 5 (high)69.31.34 (1.08–1.66)1.10 (0.84–1.44)
Comorbidity   
 No63.11.001.00
 Yes63.91.04 (0.90–1.20)1.01 (0.85–1.20)
Need for rehabilitation   
 No58.91.001.00
 Yes67.41.45 (1.26–1.65)1.40 (1.18–1.65)

Prescriptions of individual AEDs in epilepsy by specialty of physician

Lamotrigine, valproic acid, carbamazepine, oxcarbazepine, and levetiracetam were the most frequently prescribed AEDs. Neuropediatricians prescribed three of these AEDs more frequently than other specialists did (Table S1): lamotrigine (45.9% vs. 39.1%), oxcarbazepine (15.1% vs. 10.8%), and levetiracetam (14.6% vs. 7.0%), and carbamazepine less frequently (15.8% vs. 20.8%). Valproic acid was prescribed to the same extent by neuropediatricians and other specialists (41.9% vs. 43.8%).

Sociodemographic differences in access to individual AEDs

Table 3 shows adjusted odds ratios within sociodemographic subgroups for AEDs, using one stratum as the reference; (1) female patients were more likely than male patients to receive prescriptions of lamotrigine, (2) children aged 6–17 years were more likely to receive lamotrigine and less likely to receive valproic acid, oxcarbazepine, and levetiracetam as compared to children aged 1–5 years; (3) as compared to patients living in rural areas, oxcarbazepine was prescribed to a greater extent to patients living in large cities; (4) the higher the household income the greater likelihood of prescriptions of levetiracetam. The proportions of patients receiving prescriptions of the five most prescribed AEDs in different subgroups are given in Table S2 and crude odds ratios in Table S3.

Table 3.   Sociodemographic risk factors for receiving a prescription of the five most prescribed AEDs in epilepsy patients on continuous AED treatment (n = 3,631): adjusteda odds ratios
 Lamotrigine
OR (95% CI)
Valproic acid
OR (95% CI)
Carbamazepine
OR (95% CI)
Oxcarbazepine
OR (95% CI)
Levetiracetam
OR (95% CI)
  1. aSeparate odds ratios are adjusted for all the other variables in the model and county.

Sex     
 Male11111
 Female1.4 (1.2–1.6)0.9 (0.8–1.0)1.0 (0.8–1.2)0.9 (0.7–1.1)1.1 (0.9–1.3)
Age     
 1–511111
 6–121.7 (1.3–2.1)0.6 (0.5–0.8)1.3 (1.0–1.8)0.8 (0.6–1.1)0.7 (0.5–1.0)
 13–171.9 (1.5–2.3)0.6 (0.4–0.7)1.1 (0.8–1.4)0.6 (0.5–0.9)0.7 (0.5–1.0)
Parental educational level     
 Low11111
 Middle1.0 (0.9–1.2)1.1 (0.9–1.3)0.9 (0.7–1.1)1.0 (0.8–1.3)0.9 (0.7–1.2)
 High1.0 (0.8–1.2)1.0 (0.8–1.2)1.0 (0.8–1.3)1.1 (0.9–1.5)0.8 (0.6–1.1)
Parental region of birth     
 Sweden (at least one)11111
 Other country (both)0.8 (0.7–1.0)1.1 (0.9–1.3)1.3 (1.0–1.6)0.9 (0.6–1.2)1.2 (0.9–1.7)
Place of residence     
 Rural/small city11111
 Middle-sized city1.1 (0.9–1.4)0.8 (0.7–1.0)1.0 (0.8–1.3)1.5 (1.1–2.0)1.0 (0.8–1.4)
 Large city0.9 (0.7–1.2)1.0 (0.8–1.3)0.7 (0.5–1.1)2.0 (1.4–3.0)1.0 (0.7–1.6)
Parental income (quintiles)     
 1 (low)11111
 21.0 (0.8–1.2)1.0 (0.8–1.2)0.8 (0.6–1.0)1.1 (0.8–1.6)1.1 (0.8–1.6)
 31.0 (0.8–1.2)1.1 (0.9–1.4)0.9 (0.7–1.1)1.2 (0.9–1.7)1.4 (1.0–2.0)
 41.0 (0.8–1.3)1.0 (0.8–1.2)1.0 (0.8–1.3)1.2 (0.8–1.6)1.6 (1.2–2.3)
 5 (high)1.0 (0.8–1.3)1.0 (0.8–1.2)0.8 (0.6–1.0)1.0 (0.7–1.4)1.5 (1.1–2.2)
Comorbidity          
 No11111
 Yes1.1 (1.0–1.3)0.8 (0.7–0.9)1.0 (0.8–1.2)1.2 (1.0–1.5)1.1 (0.9–1.4)
Need for rehabilitation          
 No11111
 Yes1.1 (1.0–1.3)1.7 (1.5–2.0)0.7 (0.6–0.9)0.6 (0.5–0.7)1.5 (1.2–1.9)

Stratified analyses

We stratified associations between sociodemographic variables and access to prescriptions by neuropediatricians by need for rehabilitation services (Table S4) and comorbidities (Table S5).

Among patients lacking comorbidities or lacking needs of rehabilitation we found greater access to pediatric specialist among those living in middle-sized or larger cities as compared to smaller cities, whereas no such differences were found among those in need of rehabilitation or comorbid disease.

We also stratified associations between sociodemographic variables and individual AEDs by need for rehabilitation services and comorbidities (data not shown).

Access to lamotrigine prescriptions was larger in girls than in boys in the group of children lacking rehabilitation needs or without comorbidities. Within children with rehabilitation needs or comorbidities no such differences were present. Access to levetiracetam was greater in families with higher incomes as compared to families with lower incomes both in the subgroups of children with rehabilitation needs and comorbidities. No such differences were found among those without needs of rehabilitation or without comorbid disease.

Discussion

A major strength of the present study is that it is based on a nationwide large dataset. Our observations should therefore be important for policymakers in their efforts to reduce inequities in access to epilepsy care. We found that children with epilepsy on continuous AED treatment are more likely to have their prescriptions of AEDs from a neuropediatrician rather than other specialists if they live in large cities. Children living in large cities received oxcarbazepine to a greater extent than children living in rural areas. More recently licensed AEDs—lamotrigine, oxcarbazepine, and levetiracetam—were prescribed to a larger extent by neuropediatricians as compared to other physicians. Levetiracetam was prescribed more extensively to children whose parents had higher incomes than to those children with low household incomes. There is no reason to assume that the needs of children with high household incomes in terms of AED treatment are different from those with low household incomes. The cost protection means that patients’ combined expenses for any drugs during a 12-month period can reach a maximum of 1,800 Swedish kronor (approximately $300 USD).

Although our data thus indicate that sociodemographic status influences access to neuropediatricians and individual AEDs for children and adolescents with epilepsy, the inequities in access to care among children appear less pronounced than among adult Swedish epilepsy patients (Mattsson et al., 2010). Nevertheless, the results of this study are important because they show that universal coverage to medical care does not eliminate inequalities of access to health care services. The impact of giving care according to sociodemographic status and not to needs is dual. First, there is an ethical dimension. Second, treatment given according to sociodemographic belonging rather than needs may be suboptimal and can result in unfavourable health outcomes.

Our finding that 63.4% of patients received at least one prescription from a pediatrician is in line with the results obtained in the best provided European areas (Malmgren et al., 2003). In Sweden, the proportion of children receiving specialist care is higher than that of adults, 45.5% (Mattsson et al., 2010), similar to previous results in Europe (Malmgren et al., 2003). Figures reflecting the proportions of patients receiving care or prescriptions are sensitive to differences in length of observation. Children with epilepsy in our study may have had their AED prescriptions from a neuropediatrician also prior to 2006, indicating that we may underestimate the true proportion. In 2005 in Sweden, there were on average 5.0 neuropediatricians per 100,000 persons aged 0–17 years (National Board of Health and Welfare, 2007), but with an uneven geographic distribution ranging from two to five per 100,000 between the six large geographical regions. No data are available that can guide us as to whether the density of neuropediatricians is of importance to access to expert health care, but this seems likely. There is no reason to assume that the type and severity of seizures in children and adolescents is different in children living in large cities as compared to children living in rural areas.

We identified some age- and gender-related differences in prescription patterns, some of which are attributable to biologic and regulatory factors rather than sociodemographics. Girls were more likely than boys to receive prescriptions of lamotrigine, probably explained by the problems related to female reproductive function (Morrell et al., 2008). Adolescents were (1) less likely to receive prescriptions from neuropediatricians; (2) less likely to receive prescriptions of valproic acid, oxcarbazepine, and levetiracetam (the two latter more difficult to explain by biological or regulatory factors); and (3) more likely to receive lamotrigine prescriptions as compared to children aged 1–5 years. Similar findings were made in children aged 6–12 years. The highest incidence of epilepsy in childhood is during the first year of life, not covered in the present study, and many children could have been managed initially by a neuropediatrician, with treatment transferred back to nonneuropediatricians. In addition, the spectrum of epilepsy disorders (Sidenvall et al., 1996), and thus the treatment, in childhood varies with age, and there are also age limitations in the licensing of different AEDs. Therefore, it is likely that the observed age-related differences in AED prescriptions reflect age-dependent differences in epilepsy rather than true sociodemographic inequities. It is notable that access to individual AEDs does not include any measure of appropriateness.

The epilepsy prevalence of 0.56% in the present study reflects all children who had ever been given a diagnosis of epilepsy, and the prevalence of 0.24% represents only those on AED treatment (effective sample). This is similar to reported prevalence rates in children in Europe of 0.35–0.51% (Forsgren et al., 2005). The question about whether some children with epilepsy are managed by family physicians outside hospitals, and thus missed by the NPR, is relevant for the generalizability of our findings. Our prevalence rates suggest that the proportion of children with active epilepsy that are not included in our study is small and that most children with AED-treated epilepsy are included. The most severely affected children are overrepresented in the effective sample, indicating that this sample somewhat overestimates the proportion of children having access to neuropediatricians (Table 1).

This study is subject to limitations. Epilepsy coding has not been validated in the Swedish NPR, but results from the similar Danish National Hospital Registry indicate that epilepsy diagnosis had a positive predictive value of 81% (Christensen et al., 2007). In our study, diagnoses were physician based, but not formally validated. Their validity was supported by continuous AED use. Validation of coding of other diagnoses used in this study has not been performed, but a recent review of validations of disease coding in the Swedish In-patient Registry showed a positive predictive value of 85–95% (Ludvigsson et al., 2011). We were unable to identify seizure types and severity of epilepsy, characteristics that may influence access to neuropediatricians and choice of AEDs. We partially compensated for this by adjusting for important comorbidities (for example mental retardation) which are associated with severe epilepsy (Sidenvall et al., 1996). There are other factors, such as cultural differences, with potential impact on prescription patterns and access to specialists that we were not able to analyze.

We had information about the prescribed drugs that were also dispensed, but not all drugs prescribed. Failure to statistically account for clustering on neurologists is another potential limitation. If it had been possible to account for this, the reported 95% CI may in some cases have been wider.

The main strengths of this register-based study lie in its coverage of the whole population of a country, with low dropout rates. The quality of the registers is high; main diagnoses and personal identification numbers are missing in <1% (Centre for Epidemiology, 2010), and there are almost complete data on dispensed AEDs.

In summary, this large, national population based survey shows that access to neuropediatricians for children with AED-treated epilepsy depends on residency and that this access is of importance for the selection of AEDs. AED prescription patterns also varied by parental income. From a public health perspective, identifying existing disparities is an essential step in enabling the health care system to take actions to fight inequalities.

Acknowledgments

This study was sponsored by Selanders Foundation, Epilepsifonden, and the Research Fund of the Neurological Department, Uppsala University Hospital, which were involved neither in the design or conduct of the study, collection, management and analysis of the data, nor in the review of the manuscript.

Disclosure

Dr Tomson has received research grants and/or speakers honoraria from the following pharmaceutical companies: Eisai, GlaxoSmithKline, Janssen-Cilag Novartis, Sanofi-Aventis, Pfizer, and UCB-Pharma. The remaining authors have no conflicts of interest. We confirm that we have read the Journal’s position on issues involved in ethical publications and affirm that this report is consistent with those guidelines.

Ancillary