Assessing health status in Manitoba children

Acute and Chronic Conditions

ABSTRACT

Background: Numerous child health status measures have been developed, ranging from assessments of physical and mental health to activity continuums. Our objective was to report the regional distribution of physical morbidity among children in Manitoba.

Methods: Using Manitoba's population-based prescription and health care data for 1998/99, the prevalence of children with lower respiratory tract infections, four chronic conditions (asthma, cardiovascular disease, Type 1 diabetes mellitus and seizure disorders) and physical disabilities, including spina bifida and cerebral palsy, was determined for 12 Regional Health Authorities and 12 Winnipeg Community Areas, ranked by a measure of population healthiness, the premature mortality rate (PMR). Prescription rates were also reported by neighbourhood income quintile, derived from census data.

Results: Hospitalization for lower respiratory tract infection was highest in infants (6%) and increased with successive decreases in neighbourhood income or in the population healthiness of a region. On the basis of a physician diagnosis or prescription drug for asthma, 10% of school-age children had asthma. Asthma treatment rates in northern Manitoba were substantially lower than in Winnipeg. Treatment rates for cardiovascular conditions, Type I diabetes and seizure disorders approached 1% in adolescents and there were no regional differences in the distribution of these conditions. The prevalence of physical disability was highest in northern Manitoba.

Conclusion: A minority of Manitoba children suffer from chronic and serious acute health problems in childhood, but the burden of illness is not evenly distributed among children.

The 1989 Convention on the Rights of the Child proposes a broad definition of health which describes a child's right to adequate circumstances for physical, mental, spiritual, moral and social development.1 Numerous child health status measures have been developed over the years, ranging from simple scales assessing physical and mental health to activity continuums which assess the impact of illness on normal social role performance. What measures of child health should be used? The premature mortality rate has been accepted as a population health measure, but childhood mortality is an uncommon event, limiting its usefulness as a measure of health status.2 Morbidity is more common in childhood, and includes physical, developmental, psychological, emotional and behaviour problems.

Physical conditions such as chronic illness and disability impose a considerable morbidity burden for children, in terms of interference with sleep, play and school, and dependency on medications or technology.3,4 Almost 30% of children with asthma, cardiovascular conditions or diabetes are limited in their usual activities; the prevalence of activity limitation doubles in children with seizure disorders.3,5 Over 80% of children with asthma, diabetes or seizure disorders take medications, and miss on average one week of school per year. Up to 90% of children with physical disabilities are limited in their usual activities. Recurrence of acute conditions such as lower respiratory tract infections also interferes with quality of life. This paper describes the regional distribution of the following physical conditions among Manitoba's children: 1) lower respiratory tract infections, 2) four chronic conditions (asthma, cardiovascular disease, Type I diabetes mellitus and seizure disorders), and 3) physical disabilities, including spina bifida, cerebral palsy and other paralytic conditions.

METHODS Study design

This was a cross-sectional study of treatment prevalence of acute and chronic conditions in Manitoba children in 1998/99, reported by geographic area and neighbourhood income. The geographic areas used are 12 regions of Manitoba, called Regional Health Authorities (RHAs), and 12 subregions of Winnipeg, Manitoba's largest city, called Winnipeg Community Areas (Winnipeg CAs). RHAs and Winnipeg CAs were ranked by the premature mortality rate (PMR), the best single measure to represent the healthiness of a population, and its need for health care services.6,7

Study population and data sources

Administrative data were obtained on children aged 0 to 19 from the Population Health Research Data Repository (PHRDR) at the University of Manitoba. These data are anonymized, encounterbased records of Manitobans' interactions with the health care system. Four databases were used: 1) registration files, 2) records of physician reimbursement, 3) records of hospitalizations, and 4) records of prescriptions dispensed in retail pharmacies.

The registration file contains a record for every individual eligible to receive insured health services, and includes the individual's birth date, gender and geographic location. Records of physician reimbursement for medical care provided are submitted under a fee-for-service arrangement, and contain information on one patient diagnosis at the 3-digit level of the ICD-9CM classification system and physician specialty. Discharge abstracts for hospital services include information on up to 16 ICD-9-CM diagnostic codes, of which the first diagnosis is the primary diagnosis responsible for the hospital stay. Prescription records are submitted by retail pharmacies for reimbursement by provincial drug insurance plans and for drug utilization review purposes. These records contain the drug's name, identification number, dosage form, and quantity dispensed, as well as the date the drug was dispensed.

The reliability and validity of the PHRDR databases have been found to be high for describing population drug use and health care utilization for specific conditions.8,9 However, physician contact and prescription use in northern Manitoba nursing stations is incompletely recorded in the health care databases. Further, care must be taken when making inferences about disease prevalence from health care utilization data. Health care contact is less frequent among children living in rural versus urban areas.10 Further, children are more likely to use prescription medications for chronic conditions than to visit a physician on an annual basis.11 Our measure of chronic conditions also included prescription drug use to minimize bias subsequent to less frequent physician contact.

The face validity of defining chronic conditions according to diagnosis or prescription data was determined by comparing the rate of children defined on this basis, with the prevalence of children with the condition in the 1996 Manitoba sample of the National Population Health Survey (NPHS), recognizing the limitations of survey estimates of disease due to small sample size and wording of questions (i.e., ever had asthma versus current asthma). As parents may not view vascular diseases as "heart conditions," two categories of cardiovascular conditions were compared with survey data: all conditions and cardiovascular conditions excluding cerebrovascular (e.g., stroke) and vascular disorders. No comparisons with the survey data were made for the Type I diabetes case definition which had been validated previously by comparison to the Diabetes Education Resource Database.12 To assess whether regional differences in chronic conditions could be real, we compared regional variation in overall health care contact and receipt of prescription drugs with regional variation in conditionspecific treatment prevalence rates.

Study measures and analyses

The 1998/99 prevalence rate of children receiving health care was determined for: 1) lower respiratory tract infections, 2) four chronic conditions (asthma, cardiovascular disease, Type 1 diabetes mellitus and seizure disorders), and 3) physical disabilities, including spina bifida, cerebral palsy and other paralytic conditions. Hospitalization for lower respiratory tract infection included a hospitalization for pneumonia, bronchiolitis and bronchitis, and in children less than 5 years old, it also included hospitalization for asthma. 13 The treatment prevalence of asthma, cardiovascular conditions and seizure disorders was defined on the basis of at least one physician visit or hospitalization for these conditions or the receipt of at least one prescription drug to treat the condition. Children with a physical disability were identified on the basis of a physician visit or hospitalization for spina bifida, cerebral palsy or paralytic conditions from birth to 1998/99.14

The denominator was the population of children residing in Manitoba as of December 31, 1998. Age-specific and ageadjusted treatment prevalence rates were determined for each RHA and Winnipeg CA, and urban and rural neighbourhood income quintiles. Rates were suppressed where cell counts were less than 5. Children were placed into neighbourhood income quintiles, RHAs, and Winnipeg CAs, according to the postal code of their place of residence. Rural and urban neighbourhood income quintiles were created from Statistics Canada Census 1996 by aggregating household income data to the census enumeration area and ranking neighbourhood income quintiles from 20% of the population residing in the lowest income neighbourhoods to 20% of the population residing in the highest income neighbourhoods.15-17

Using t-test methodology developed by Carriere and Roos,18 treatment prevalence rates for individual RHAs and Winnipeg CAs were compared, at the 99% level of confidence, to the total Manitoba child population, Winnipeg, Non-Winnipeg (all children residing outside of Winnipeg), rural south (children residing in rural regions below the 53rd parallel) and north (children residing in regions above the 53rd parallel). Visual inspections of the regional rates, ranked according to PMR (lowest to highest), were performed to provide insight into the relationship with the healthiness of the population within a region (PMR). To further assess the association with PMR, treatment prevalence rates for regions were correlated with region PMR, using Spearman's rank correlation. Trends in prevalence rates across income quintiles were assessed with the Cochran-Armitage test for trend at the 95% level of confidence.

RESULTS

Validation of chronic conditions

Age-specific asthma treatment rates fell within the 95% confidence intervals for asthma prevalence reported in the NPHS for children less than 15 years of age, indicating similarity in prevalence (Table I). The NPHS documented a higher prevalence of adolescents with asthma than that derived from health care administrative data, suggesting under-representation by health care administrative data or survey over-reporting of asthma which was no longer current.19 The treatment rates of the other conditions were similar to survey rates, but because of their low prevalence, survey rates were unreliable.

Fewer children in Northern Manitoba had health care use or received prescription drugs than children living in Winnipeg. The ratio of the prevalence of any health care for Winnipeg children (86.3%) versus northern Manitoba children (67.6%) was 1.1. The ratio of the treatment prevalence for cardiovascular conditions between these regions was similar (Winnipeg/North ratio=1.3). However, the Winnipeg-toNorth treatment prevalence ratio for asthma was 2.7.

Treatment prevalence rates

Eighty percent of Manitoba children saw a physician or were hospitalized in 1998/99 and 59% received at least one prescription drug. Defining health care contact on the basis of use of physician services, hospitalization or the receipt of a prescription drug, 83% of children received health care treatment.

1) Lower respiratory tract infection. In 1998/99, less than 1% of all children were hospitalized for a lower respiratory tract infection (LRI). The prevalence rate of hospitalization for LRI was highest in children less than one year of age (6% of infants), but declined dramatically by the age of five. The prevalence rate of LRI hospitalization in infants was highly correlated with the PMR (the healthiness) of the RHA populations (Spearman rank correlation =0.66, p<0.02). The rate in the NorMan, Burntwood and Parkland RHAs was significantly higher than the southern rural and the provincial average (Figure 1). Hospitalization for LRI was also highly correlated with the healthiness of populations in the Winnipeg CAs (Spearman rank correlation=0.61, p<0.04) (Figure 2). Infants living in the Point Douglas area were significantly more likely to be hospitalized for LRI than all Winnipeg infants combined (Figure 2). Of note, the prevalence rate of LRI hospitalization in the northern RHAs (150 per 1,000 infants) was 10-fold greater than the rate for infants living in the Winnipeg communities with the healthiest populations.

Hospitalization for LRI represents only the "tip of the iceberg" for all LRI morbidity as shown when physician contacts for LRI are combined with hospitalization rates (Figure 1). However, under-reporting of physician contacts in nursing stations makes this a less useful measure for comparisons across RHAs. Furthermore, there may be variation in LRI hospitalization by income neighbourhoods within RHAs and Winnipeg CAs. We observed increases in LRI hospitalization with decreasing neighbourhood income (Figure 3); this gradient effect was substantially steeper in rural areas (Cochran-Armitage trend test, p<0.0001).

2) Asthma. Ten percent of Manitoba's school-age children (5-19 years old) had either a physician diagnosis of asthma or a prescription for an asthma drug in 1998/99. The asthma treatment prevalence declined with age, from 11.5% in children 5-9 years to 8% in adolescents.

Asthma treatment rates were highest in Winnipeg (and Brandon in younger children), followed by southern rural areas, and then by the North (Figure 4). Among southern rural RHAs, the asthma treatment prevalence rate was significantly higher in the Interlake region than the southern rural average. Within Winnipeg, most asthma treatment rates did not differ by Winnipeg CA from the Winnipeg average, but a significantly higher rate was observed in the St. James-Assiniboine area for children 10 years of age and older.

3) Other chronic conditions. As documented for 1996/97 (Table I), the treatment prevalence for other chronic conditions in 1998/99 increased with age and was low in children less than 15 years old. On the basis of a physician visit, hospitalization or prescription drug, 1.2% of children, aged 15-19 years, had a cardiovascular condition, 0.7% had a seizure disorder, and 0.3% had Type I diabetes mellitus. In this age group, no statistically significant differences for the treatment prevalence of cardiovascular conditions, seizure disorders and type I diabetes were observed among RHAs or Winnipeg CAs. Age comparisons of the treatment prevalence for cardiovascular conditions (most frequent diagnosis was hypertension) by neighbourhood income documented the beginnings of a graded association with income (Figure 5). Higher cardiovascular condition rates were observed with decreases in neighbourhood income in urban areas among older children [Cochran-Armitage trend test for 5-9 year olds (NS), for 10-14 year olds (p<0.05), for 15-19 year olds (p<0.10)].

4) Physical disability (spina bifida, cerebral palsy and other paralytic conditions). In 1998/99, less than 1% of children were found to have spina bifida, cerebral palsy or paralytic conditions, determined on the basis of their health care utilization records from birth. Figure 6 shows age-standardized disability rates by Winnipeg, Brandon, and northern and southern Manitoba. Winnipeg and northern regions were more likely to have children with these disabilities, but the only statistically significant difference was between northern versus southern rural rates of cerebral palsy/paralytic conditions.

DISCUSSION

Over 80% of Manitoba children received health care in 1998/99. However, health care for the acute and chronic conditions selected for study was not common. Hospitalization of children for respiratory conditions has declined over the past two decades;" less than 1% of all Manitoba children were hospitalized for LRI. However, LRI hospitalization rates were highest in infants and are a significant cause of mortality in this age group.21 Infants living in low-income regions or the least healthy regions in Manitoba, as indicated by the PMR, were much more likely to be hospitalized for LRI. Similarly, higher LRI hospitalization rates among young children have been reported in US geographic areas characterized by higher rates of poverty.13 A variety of risk factors for LRI, such as household crowding, smoking and lower breastfeeding rates, have been associated with living in a low-income household.22 Manitoba regions with the highest LRI hospitalization rates also had the highest rates of these risk factors.23

Asthma was the most common chronic condition in childhood. On the basis of physician diagnosis or prescription use for asthma, 10% of school-age children had asthma in 1998/99, similar to the prevalence reported by the recent Canadian Student Lung Health Survey.11 Not all children with asthma symptoms continue to have asthma as they grow older.19 In our study, the treatment prevalence for asthma declined with age, compatible with this "growing out of asthma" phenomenon. Much lower asthma treatment rates were observed in northern Manitoba. While this finding may be a function of missing data on physician visits and prescriptions dispensed in nursing stations, relative differences in asthma treatment prevalence between Winnipeg and northern Manitoba were in excess of differences for health care contact between the two regions. Further, an asthma treatment definition which included prescription drug use potentially diminished bias from urban-rural differences in physician contact.10,11 Our observations of lower asthma treatment rates in northern and southern rural regions do have some biological plausibility. First, a lower prevalence of asthma has been reported in Aboriginal children living in rural Manitoba and Australia.24,25 Second is the cleanliness hypothesis which has been proposed to explain geographical differences in asthma prevalence.26 The hypothesis states that children exposed to "germs" at an earlier age are much less likely to get asthma. We observed that northern and southern regions were also more likely than Winnipeg to have infants with lower respiratory tract infections.

Other chronic conditions in childhood were much less common than asthma, and their prevalence increased with age.27 In our research, treatment rates for cardiovascular conditions, Type I diabetes and seizure disorders approached 1% in adolescents and there were no regional differences in the distribution of these conditions. We reported only on the prevalence of Type I diabetes in children, but there are regional differences in the prevalence of Type II diabetes, a disease which is reaching epidemic proportions in young Aboriginals.28 Higher cardiovascular condition rates were observed with decreases in neighbourhood income in urban areas. This phenomenon has been documented by others, and has been attributed to the differential acquisition in children of risk factors, such as sedentary lifestyle and smoking, by household income.29

Physical disability can be the outcome of the prenatal period or of events later in childhood secondary to injury. The prevalence of physical disability was higher in the North, which we postulate is the outcome of higher rates of injury.30 These disabilities cause a major interference with the activities of these children, significant consumption of health care services, and a high level of dependence on families for personal care.31

In summary, a minority of Manitoba children suffer from chronic and serious acute health problems in childhood, but the burden of illness is not evenly distributed among children. Acute conditions such as respiratory tract infections, and physical disability are more common in Northern Manitoba children, but asthma is more common in urban children. Our validity assessments suggest that these regional differences are real. However, we have presented data on physical health in Manitoba children, recognizing that physical health represents only one dimension of child health. Other dimensions, such as mental and emotional health, are equally as important and may be more pervasive in children.

[Reference]

REFERENCES

[Reference]

1. Pal DK. Quality of life assessment in children: A review of conceptual and methodologic issues in multidimensional health status measures. J Epidemiol Community Health 1996;50:391-96.

2. Szilagyi PG, Schor EL. The health of children. Health Serv Res 1998;33:1001-39.

3. Newacheck PW, Halfon N. Prevalence and impact of disabling chronic conditions in childhood. Am Public Health 1998;88:610-17.

4. Stein REK, Bauman LJ, Westbrook LE, Coupey SM, Ireys HT. Framework for identifying children who have chronic conditions: The case for a new definition./Pediatr 1993;122:342-47.

5. Newacheck PW, Taylor WR. Childhood chronic illness: Prevalence, severity, and impact. Am J Public Health 1992;82:364-71.

6. Carstairs V, Morris R. Deprivation and Health in Scotland. Aberdeen, Scotland: Aberdeen University Press, 1991.

7. Eyles J, Birch S, Chambers S, Hurley J, Hutchison B. A needs-based methodology for allocating health care resources in Ontario: Development and an application. Soc Sci Med 1991;33(4):489-500.

8. Kozyrskyj A, Mustard CA. Validation of an electronic, population-based prescription database. Ann Pharmacother 1998;32:1152-57.

9. Robinson JR, Young TK, Roos LL, Gelskey DE. Estimating the burden of disease: Comparing administrative data and self-reports. Med Care 1997;35:932-47.

[Reference]

10. Kozyrskyj AL, Mustard CA, Derksen S. Considering the Health Care Needs of Children Living in Households Receiving Income Assistance in Manitoba, 2000. Winnipeg: Manitoba Centre for Health Policy.

11. Health Protection Branch, Laboratory Centre for Disease Control, Health Canada. Childhood Asthma in Sentinel Health Units: Findings of the Student Lung Health Survey 1995-1996. Ottawa, 1998.

12. Blanchard J, Dean H, Anderson K, Wajda A, Ludwig S, Depew N. Incidence and prevalence of diabetes in children aged 0-14 years in Manitoba, Canada, 1985-1993. Diabetes Care 1997;20:51215.

13. McConnochie KM, Roghmann KJ, Liptak GS. Hospitalization for lower respiratory tract illness in infants: Variation in rates among counties in New York State and areas within Monroe County.J Pediatr 1995;126:220-29.

14. Palfrey JS, Singer JD, Walker DK, Butler JA. Early identification of children's special needs: A study in five metropolitan communities. JPediatr 1987;111:651-59.

15. Mustard CA, Roos NP. The relationship of prenatal care and pregnancy complications to birthweight in Winnipeg, Canada. Am JPublic Health 1992;82:1119-26.

16. Wilkins R. Use of postal codes and addresses in the analysis of health data. Health Rep 1993;5:157-77.

17. Mustard CA, Derksen S, Berthelot J-M, Wolfson W. Assessing ecologic proxies for household income: A comparison of household and neighbourhood level income measures in the study of population health status. Health and Place 1999;5:157-71.

18. Carriere KC, Roos LL. A method of comparison for standardized rates of low-incidence events. Med Care 1997;35 (1):57-69.

19. Barbee RA, Murphy S. The natural history of asthma. J Allergy Clin Immunol 1998;102:S65572.

20. Hodge MJ, Dougherty GE, Pless IB. Pediatric mortality and hospital use in Canada and the United States, 1971 through 1987. Am J Public Health 1995;85:1276-79.

[Reference]

21. Martens PJ, Derksen S, Mayer T, Walld R. Being born in Manitoba: A look at perinatal health issues. Can J Public Health 2002;93(Suppl. 2):533-S38.

22. Nafstad P, Jaakkola JJK, Hagen JA, Botten G, Kongerud J. Breastfeeding, maternal smoking and lower respiratory tract infections. Eur Respir J 1996;9:2623-29.

23. Martens PJ, Derksen S. A matter of life and death for Manitoba's children: An overview of birth rates and mortality rates. Can J Public Health 2002;93(Suppl. 2):521-S26.

24. Pasterkamp H, Moffatt M, Alkrinawi S, Holbrow J. Airway responsiveness in school children of the Canadian First Nations [abstract]. International Conference, May 1996. New Orleans.

25. Whybourne AM, Lesnikowski CL, Ruben AR, Walker AC. Low rates of hospitalization for asthma among Aboriginal children compared to nonAboriginal children of the Top End of the Northern Territory. J Paediatr Child Health 1999;35:438-41.

26. Cookson W, Moffatt M. Asthma: An epidemic in the absence of infection? Science 1997;275:4142.

27. Mustard C, Derksen S, Berthelot JM, Wolfson M, Roos LL. Age-specific education and income gradients in morbidity and mortality in a Canadian province. Soc Sci Med 1997;45:383-97.

28. Dean H. NIDDM-Y in First Nation children in Canada. Clin Pediatr 1998;37:89-96.

29. Leino M, Porkka KV, Raitakari OT, Laitinen S, Taimela S, Viikari JS. Influence of parental occupation on coronary heart disease risk factors in children. The Cardiovascular Risk in Young Finns Study. IntJEpidemiol 1996;25:1189-95.

30. MacKenzie EJ. Epidemiology of injuries: Current trends and future challenges. Epidemiol Rev 2000;22(1):112-19.

31. Blum RW, Resnick MD, Nelson R, St Germaine A. Family and peer issues among adolescents with spina bifida and cerebral palsy. Pediatrics 1991;88:280-85.

[Reference]

RESUME

[Reference]

Contexte : Il existe de nombreuses mesures de l'etat de sante des enfants; elles vont des evaluations de la sante physique et mentale aux continuums de I'activite physique. On presence ici la repartition regionale des troubles physiques chez les enfants manitobains.

Methode : A l'aide des donnees representatives du Manitoba sur la prescription cle medicaments et les soins de sante (1998-1999), nous avons determine la prevalence des enfants presentant des infections des voies respiratoires inferieures, quatre etats chroniques (asthme, maladie cardiovasculaire, diabete de type 1, troubles epileptiques) ou des deficiences physiques comme le spina-bifida et l'infirmite motrice cerebrale dans 12 ORS et 12 CR cle Winnipeg, classes selon une mesure de l'etat de sante de la population (le TMP). Les taux de prescription de medicaments sur ordonnance sont egalement presentes par quintile de revenu des quartiers (selon les donnees du recensement).

[Reference]

Resultats: Le plus haut taux d'hospitalisation pour infection des voles respiratoires inferieures a ete mesure chez les nourrissons (6 %) et augmentait avec les diminutions successives du revenu du quartier ou de I'etat de sante de la population de la sous-region. D'apres les diagnostics des medecins et les medicaments prescrits contre I'asthme, 10 % des enfants d'Age scolaire etaient asthmatiques. Les taux de traitement de I'asthme dans le Nord du Manitoba etaient sensiblement plus faibles qu'a Winnipeg. Les taux de traitement des maladies card iovascu lai res, du diabete de type 1 et des troubles epileptiques atteignaient pres de 1 % chez les adolescents; on n'observe aucun ecart regional dans la repartition de ces troubles. C'est dans le Nord du Manitoba que la prevalence des deficiences physiques etait la plus elevee.

Conclusion: Une minorite d'enfants manitobains souffre de problemes de sante chroniques et graves durant l'enfance, mais le fardeau de la maladie West pas equitablement reparti entre les enfants.

[Author Affiliation]

1. Manitoba Centre for Health Policy, Department of Community Health Sciences, Faculty ot Medicine, University of Manitoba, Winnipeg, MB

2. Department Pediatrics and Child Health, Faculty of Medicine, University of Manitoba

[Author Affiliation]

Correspondence: Dr. Anita Kozyrskyj, Manitoba Centre for Health Policy, Department ot Community Health Sciences, Faculty of Medicine, University of Manitoba, 4th Floor Brodie Centre, Room 408 727 McDermot Avenue, Winnipeg, MB R3E 3P5. The full report "Assessing the Health of Children in Manitoba: A Population-Based Study" on which this article is based is available from the Manitoba Centre for Health Policy at the above address or online at: http://www.u manitoba.ca/centres/mchp/reports.htm

Sources of support: This work was supported as part of a project on Child Health in Manitoba, one of several projects undertaken each year by the Manitoba Centre for Health Policy under contract to Manitoba Health. The results and conclusions are those of the authors and no official endorsement by Manitoba Health was intended or should be inferred.