Shared help seeking behaviour within families: a retrospective cohort
http://www.100md.com
《英国医生杂志》
1 NIVEL (Netherlands Institute for Health Services Research), PO Box 1568, 3800 BN Utrecht, Netherlands, 2 Department of General Practice, UMCN St Radboud, PO Box 9101, 6500 HB Nijmegen, Netherlands
Correspondence to: M Cardol m.cardol@nivel.nl
Objective To examine the extent to which the family influences individual use of general practitioner care.
Design Retrospective cohort study of all consultations in one calendar year. Multilevel modelling was used to analyse contact frequencies of individuals within families within practices.
Setting General practice in the Netherlands.
Participants 42 262 families with children aged 2-21 years registered in 96 practices.
Main outcome measures Family influence on individual frequency of contact with general practice and correlation in frequency of contacts between parents and children.
Results After correction for patients' age and sex, analysis of siblings indicates that 22% of the variance in frequencies of contact can be ascribed to influence of the family. This means that contact frequencies of family members within families resemble each other, whereas differences in contact frequencies exist between families. Almost 6% of the variance refers to differences between practices and 73% of the variance refers to individual differences. The strongest correlations were found between mothers and children and between children.
Conclusions The extent of shared help seeking behaviour within families has considerable implications in the context of the practice.
Efforts by general practitioners to promote good health or to influence consultation patterns may conflict with their patients' family habits and attitudes at home. In general practice the family was long considered as a starting point for treatment. Families are important social contexts within which illness occurs, lingers, or resolves. Families share the same lifestyle and home environment, and they share beliefs and behaviours relating to illness and health, thereby influencing each other's use of medical care.1–6 Most publications emphasise the importance of the mother's role in relation to children's health, with less research on the contribution of fathers.1 3 The propensity of some families to use more health services than others may be attributed to predisposing factors such as family composition, health beliefs, and social structure.
Family members show similar help seeking behaviour with regard to morbidity over time7 and the relation between morbidity and attendance,8 while consultation patterns within the family are even transferred to succeeding generations.2 In specific terms, consulting a general practitioner for minor ailments, such as a headache or abdominal pain, can affect the replication of consulting patterns from parents to offspring because freedom of choice to consult a general practitioner is most clear in the case of minor ailments.9 It is therefore important to consider patients' social contexts with a view to prevention, diagnosis, and treatment in general practice.
Research on the role of the family, however, mostly dates from the 1970s and 1980s and the family is scarcely mentioned thereafter, let alone used as a unit in analysis. Have individualisation theory, evidence based medicine, and a patient centred approach suppressed the ideas of family medicine?
The individualisation hypothesis suggests that attitudes and behaviour are increasingly based on personal choice and are less dependent on tradition and social connections.10 This has reduced the impact of the family: families are less cohesive and members are more autonomous, while parenting has become less controlling than some decades ago.11 Children now have a more active role in their interaction with adults and understand more about concepts of health and illness than presumed.12 13 Evidence based medicine is also characterised by an individualistic approach, and this differs from the systems theory, in which family science is rooted.14 Furthermore, the structure of families has changed, and this may well have moderated the influence of families on health beliefs and consequently on consultation patterns. One parent families are now more common in the Netherlands. In two parent families, one parent more often is not the biological parent, and both parents more often have paid jobs. On the other hand, family influence on consultation rates may have increased because in smaller families each child gets more attention, and only 20% of families now have more than two children.15
We examined the extent to which families continue to influence individual use of general practitioner care.
Method
We used data from the second Dutch national survey of general practice. The survey recorded all consultations in 2001 for 104 general practices in the Netherlands, comprising 195 general practitioners serving 385 461 patients.16
We selected for analysis families with one or more children aged 21 years to exclude elderly parents who live with their adult children. The minimum age of the children was set at 2 years because children below this age also attend baby or child health clinics.
For all family members we selected only those consultations for new problems, as these are the contacts in which the initiative of the patient is most clear. We excluded eight practices from the analysis, mainly because of technical problems with registration. Additional analysis showed that excluded practices did not differ from the included practices in terms of practice size, practice type, and degree of urbanisation.
Relation between frequencies of first contacts and patients' age and sex, corrected for influences of family and practice
We considered context in the analysis because this influences people's care needs, what they want to do, and what they can do. In addition to the individual level, we considered two kinds of contexts: the family and the practice. Multilevel analysis enabled us to analyse the impact of the family on individual frequency of contacts while also considering another important context related to contact frequencies—that is, the practice. Multilevel analysis extends single level regression analysis to settings with hierarchical data. We calculated the variance in frequencies of contacts due to differences between individuals, differences between families, and differences between practices.17 Accordingly, the total variance is the sum of the variance on three levels: individual, family, and practice.
We identified the family impact by the amount of variance in the frequency of first contacts with the general practice at family level. Greater impact of family background should result in more variance at family level, indicating more resemblance between family members with respect to frequency of contacts. At the same time, shared help seeking behaviour within families indicates more differences between families. We calculated the variance when all family members were included in the analysis and then carried out a second analysis in which only siblings were included. We further described the family impact by correlation coefficients to evaluate the magnitude of the family impact in relationships between parents and children.
Frequencies of contacts are not normally distributed; they are discrete rather than continuous and usually skewed. Therefore, we used a Poisson distribution in the multilevel analyses (a linear model was used for the estimates of the variances on the three levels because correlations within classes cannot be estimated correctly on the individual level when applying a Poisson distribution).18
Frequency of contacts also differs by age and sex (figure). To capture the non-linear relation between age, sex, and frequency of contacts, we modelled age as a separate effect for four groups (father, mother, son, daughter).
Results
Table 1 shows the composition of the study population. We included over 42 000 families with children aged 2-21 years living at home (160 926 people in total). Almost 18% were one parent families, and about 1% comprised parents of the same sex, three generations in one home, or a compound family structure. The average number of children per family was two, with a maximum of 11.
Table 1 Composition of study population (n=160 926)
Corrected for patients' age and sex, 6% of the variance in frequency of contacts clustered on practice level, suggesting that similarities in contact frequencies within families differ between practices. About 18% of the variance can be ascribed to family influence and 76% to individual differences.
When we excluded the parents from the analysis, the family accounted for about 22% of the variance in contact frequencies, the practice for 6%, and individual differences between siblings for 73%. This means that more than a fifth of the variance in contact frequencies relates to shared help seeking behaviour within families.
Table 2 shows the correlations between family members according to sex. The strongest correlations were between contact frequencies for mothers and children and between children. The association between contact frequencies for fathers and children was about the same as the association between parents: somewhat lower but still substantial. The association between parents shows that resemblances in contact frequencies between family members cannot be ascribed to genetic factors alone.
Table 2 Correlations in frequencies of first contacts with general practice within families, according to sex, corrected for age and sex (n=160 926)
Discussion
Mechanic D. The influence of mothers on their children's health attitudes and behavior. Pediatrics 1964;33: 444-53.
Huijgen FJA. Family medicine; the medical life history of families. Nijmegen: Dekker and Van de Vegt, 1978.
Litman TL. The family as a basic unit in health and medical care: a social-behavioral overview. Soc Sci Med 1974;8: 495-519.
Schor E, Starfield B, Stidley C, Hankin J. Family health. Med Care 1987;25: 616-26.
Wilcox-Gok VL. Sibling data and the family background influence on child health. Med Care 1983;21: 630-8.
Bosch WJHM van de. Epidemiologische aspecten van morbiditeit bij kinderen . Nijmegen: University of Nijmegen, 1992.
Hippisley-Cox J, Coupland C, Pringle M, Crown N, Hammersley V. Married couples' risk of same disease: cross sectional study. BMJ 2002;325: 636-40.
Starfield B, Katz H, Gabriel A, Livingstone G, Benson P, Hankin J, et al. Morbidity in childhood: a longitudinal view. N Engl J Med 1984;310: 824-9.
Stewart P, O'Dowd T. Clinically inexplicable frequent attenders in general practice. Br J Gen Pract 2002;52: 1000-1.
Ester P, Halman L, Moor R de. Value shifts in western societies. In: Ester P, Halman L, de Moor R, eds. The individualizing society: value change in Europe and North America. Tilburg, Netherlands: Tilburg University Press, 1993: 1-20.
Swaan A de. In care of the state: health care, education, and welfare in Europe and the USA in the modern area. Cambridge: Polity Press, 1988.
Elbers E, Maier R, Hoekstra T, Hoogsteder M. Internalization and adult-child interaction. Learn Instruct 1992;2: 101-18.
Tates K, Meeuwesen L. "Let mum have her say": turntaking in doctor-parent-child communication. Patient Educ Couns 2000;40: 151-62.
Campbell TL, Culpepper L. Family medicine. In: Jones R, Britten N, Culpepper L, Gass D, Grol R, Mant D, et al. Oxford textbook of primary medical care. Vol 1. Oxford: Oxford University Press, 2004: 299-309.
Alders M. Demografie van gezinnen . Bevolkingstrends: Statistisch kwartaalblad over de demografie van Nederland 2003:51: 31-4.
Westert GP, Schellevis FG, de Bakker DH, Groenewegen PP, Bensing JM, van der Zee J. Monitoring health inequalities through general practice: the second Dutch national survey of general practice. Eur J Public Health 2005 (in press).
Leyland AH, Groenewegen PP. Multilevel modelling and public health policy. Scand J Public Health 2003;31: 267-74.
Snijders TAB, Bosker RJ. Multilevel analysis: an introduction to basic and advanced multilevel modelling. London: Sage, 1999: 234-8.
Monden C. Education, inequality and health . Nijmegen: University of Nijmegen, 2003.
Cornford CS, Cornford HM. "I'm only here because of my family." A study of lay referral networks. Br J Gen Pract 1999;49: 617-20.
Cunningham M. The influence of parental attitudes and behaviors on children's attitudes toward gender and household labor in early adulthood. J Marriage Fam 2001;63: 111-22.
Whiteman SD, McHale SM, Crouter AC. What parents learn from experience: the first child as a first draft? J Marriage Fam 2003;65: 608-21.
Parcel TL, Dufur MJ. Capital at home and at school: effects on child social adjustment. J Marriage Fam 2001;63: 32-47.
Launer J, Lindsey C. Training for systematic general practice: a new approach from the Tavistock clinic. Br J Gen Pract 1997;47: 453-6.(M Cardol, researcher1, P P Groenewegen, )
Correspondence to: M Cardol m.cardol@nivel.nl
Objective To examine the extent to which the family influences individual use of general practitioner care.
Design Retrospective cohort study of all consultations in one calendar year. Multilevel modelling was used to analyse contact frequencies of individuals within families within practices.
Setting General practice in the Netherlands.
Participants 42 262 families with children aged 2-21 years registered in 96 practices.
Main outcome measures Family influence on individual frequency of contact with general practice and correlation in frequency of contacts between parents and children.
Results After correction for patients' age and sex, analysis of siblings indicates that 22% of the variance in frequencies of contact can be ascribed to influence of the family. This means that contact frequencies of family members within families resemble each other, whereas differences in contact frequencies exist between families. Almost 6% of the variance refers to differences between practices and 73% of the variance refers to individual differences. The strongest correlations were found between mothers and children and between children.
Conclusions The extent of shared help seeking behaviour within families has considerable implications in the context of the practice.
Efforts by general practitioners to promote good health or to influence consultation patterns may conflict with their patients' family habits and attitudes at home. In general practice the family was long considered as a starting point for treatment. Families are important social contexts within which illness occurs, lingers, or resolves. Families share the same lifestyle and home environment, and they share beliefs and behaviours relating to illness and health, thereby influencing each other's use of medical care.1–6 Most publications emphasise the importance of the mother's role in relation to children's health, with less research on the contribution of fathers.1 3 The propensity of some families to use more health services than others may be attributed to predisposing factors such as family composition, health beliefs, and social structure.
Family members show similar help seeking behaviour with regard to morbidity over time7 and the relation between morbidity and attendance,8 while consultation patterns within the family are even transferred to succeeding generations.2 In specific terms, consulting a general practitioner for minor ailments, such as a headache or abdominal pain, can affect the replication of consulting patterns from parents to offspring because freedom of choice to consult a general practitioner is most clear in the case of minor ailments.9 It is therefore important to consider patients' social contexts with a view to prevention, diagnosis, and treatment in general practice.
Research on the role of the family, however, mostly dates from the 1970s and 1980s and the family is scarcely mentioned thereafter, let alone used as a unit in analysis. Have individualisation theory, evidence based medicine, and a patient centred approach suppressed the ideas of family medicine?
The individualisation hypothesis suggests that attitudes and behaviour are increasingly based on personal choice and are less dependent on tradition and social connections.10 This has reduced the impact of the family: families are less cohesive and members are more autonomous, while parenting has become less controlling than some decades ago.11 Children now have a more active role in their interaction with adults and understand more about concepts of health and illness than presumed.12 13 Evidence based medicine is also characterised by an individualistic approach, and this differs from the systems theory, in which family science is rooted.14 Furthermore, the structure of families has changed, and this may well have moderated the influence of families on health beliefs and consequently on consultation patterns. One parent families are now more common in the Netherlands. In two parent families, one parent more often is not the biological parent, and both parents more often have paid jobs. On the other hand, family influence on consultation rates may have increased because in smaller families each child gets more attention, and only 20% of families now have more than two children.15
We examined the extent to which families continue to influence individual use of general practitioner care.
Method
We used data from the second Dutch national survey of general practice. The survey recorded all consultations in 2001 for 104 general practices in the Netherlands, comprising 195 general practitioners serving 385 461 patients.16
We selected for analysis families with one or more children aged 21 years to exclude elderly parents who live with their adult children. The minimum age of the children was set at 2 years because children below this age also attend baby or child health clinics.
For all family members we selected only those consultations for new problems, as these are the contacts in which the initiative of the patient is most clear. We excluded eight practices from the analysis, mainly because of technical problems with registration. Additional analysis showed that excluded practices did not differ from the included practices in terms of practice size, practice type, and degree of urbanisation.
Relation between frequencies of first contacts and patients' age and sex, corrected for influences of family and practice
We considered context in the analysis because this influences people's care needs, what they want to do, and what they can do. In addition to the individual level, we considered two kinds of contexts: the family and the practice. Multilevel analysis enabled us to analyse the impact of the family on individual frequency of contacts while also considering another important context related to contact frequencies—that is, the practice. Multilevel analysis extends single level regression analysis to settings with hierarchical data. We calculated the variance in frequencies of contacts due to differences between individuals, differences between families, and differences between practices.17 Accordingly, the total variance is the sum of the variance on three levels: individual, family, and practice.
We identified the family impact by the amount of variance in the frequency of first contacts with the general practice at family level. Greater impact of family background should result in more variance at family level, indicating more resemblance between family members with respect to frequency of contacts. At the same time, shared help seeking behaviour within families indicates more differences between families. We calculated the variance when all family members were included in the analysis and then carried out a second analysis in which only siblings were included. We further described the family impact by correlation coefficients to evaluate the magnitude of the family impact in relationships between parents and children.
Frequencies of contacts are not normally distributed; they are discrete rather than continuous and usually skewed. Therefore, we used a Poisson distribution in the multilevel analyses (a linear model was used for the estimates of the variances on the three levels because correlations within classes cannot be estimated correctly on the individual level when applying a Poisson distribution).18
Frequency of contacts also differs by age and sex (figure). To capture the non-linear relation between age, sex, and frequency of contacts, we modelled age as a separate effect for four groups (father, mother, son, daughter).
Results
Table 1 shows the composition of the study population. We included over 42 000 families with children aged 2-21 years living at home (160 926 people in total). Almost 18% were one parent families, and about 1% comprised parents of the same sex, three generations in one home, or a compound family structure. The average number of children per family was two, with a maximum of 11.
Table 1 Composition of study population (n=160 926)
Corrected for patients' age and sex, 6% of the variance in frequency of contacts clustered on practice level, suggesting that similarities in contact frequencies within families differ between practices. About 18% of the variance can be ascribed to family influence and 76% to individual differences.
When we excluded the parents from the analysis, the family accounted for about 22% of the variance in contact frequencies, the practice for 6%, and individual differences between siblings for 73%. This means that more than a fifth of the variance in contact frequencies relates to shared help seeking behaviour within families.
Table 2 shows the correlations between family members according to sex. The strongest correlations were between contact frequencies for mothers and children and between children. The association between contact frequencies for fathers and children was about the same as the association between parents: somewhat lower but still substantial. The association between parents shows that resemblances in contact frequencies between family members cannot be ascribed to genetic factors alone.
Table 2 Correlations in frequencies of first contacts with general practice within families, according to sex, corrected for age and sex (n=160 926)
Discussion
Mechanic D. The influence of mothers on their children's health attitudes and behavior. Pediatrics 1964;33: 444-53.
Huijgen FJA. Family medicine; the medical life history of families. Nijmegen: Dekker and Van de Vegt, 1978.
Litman TL. The family as a basic unit in health and medical care: a social-behavioral overview. Soc Sci Med 1974;8: 495-519.
Schor E, Starfield B, Stidley C, Hankin J. Family health. Med Care 1987;25: 616-26.
Wilcox-Gok VL. Sibling data and the family background influence on child health. Med Care 1983;21: 630-8.
Bosch WJHM van de. Epidemiologische aspecten van morbiditeit bij kinderen . Nijmegen: University of Nijmegen, 1992.
Hippisley-Cox J, Coupland C, Pringle M, Crown N, Hammersley V. Married couples' risk of same disease: cross sectional study. BMJ 2002;325: 636-40.
Starfield B, Katz H, Gabriel A, Livingstone G, Benson P, Hankin J, et al. Morbidity in childhood: a longitudinal view. N Engl J Med 1984;310: 824-9.
Stewart P, O'Dowd T. Clinically inexplicable frequent attenders in general practice. Br J Gen Pract 2002;52: 1000-1.
Ester P, Halman L, Moor R de. Value shifts in western societies. In: Ester P, Halman L, de Moor R, eds. The individualizing society: value change in Europe and North America. Tilburg, Netherlands: Tilburg University Press, 1993: 1-20.
Swaan A de. In care of the state: health care, education, and welfare in Europe and the USA in the modern area. Cambridge: Polity Press, 1988.
Elbers E, Maier R, Hoekstra T, Hoogsteder M. Internalization and adult-child interaction. Learn Instruct 1992;2: 101-18.
Tates K, Meeuwesen L. "Let mum have her say": turntaking in doctor-parent-child communication. Patient Educ Couns 2000;40: 151-62.
Campbell TL, Culpepper L. Family medicine. In: Jones R, Britten N, Culpepper L, Gass D, Grol R, Mant D, et al. Oxford textbook of primary medical care. Vol 1. Oxford: Oxford University Press, 2004: 299-309.
Alders M. Demografie van gezinnen . Bevolkingstrends: Statistisch kwartaalblad over de demografie van Nederland 2003:51: 31-4.
Westert GP, Schellevis FG, de Bakker DH, Groenewegen PP, Bensing JM, van der Zee J. Monitoring health inequalities through general practice: the second Dutch national survey of general practice. Eur J Public Health 2005 (in press).
Leyland AH, Groenewegen PP. Multilevel modelling and public health policy. Scand J Public Health 2003;31: 267-74.
Snijders TAB, Bosker RJ. Multilevel analysis: an introduction to basic and advanced multilevel modelling. London: Sage, 1999: 234-8.
Monden C. Education, inequality and health . Nijmegen: University of Nijmegen, 2003.
Cornford CS, Cornford HM. "I'm only here because of my family." A study of lay referral networks. Br J Gen Pract 1999;49: 617-20.
Cunningham M. The influence of parental attitudes and behaviors on children's attitudes toward gender and household labor in early adulthood. J Marriage Fam 2001;63: 111-22.
Whiteman SD, McHale SM, Crouter AC. What parents learn from experience: the first child as a first draft? J Marriage Fam 2003;65: 608-21.
Parcel TL, Dufur MJ. Capital at home and at school: effects on child social adjustment. J Marriage Fam 2001;63: 32-47.
Launer J, Lindsey C. Training for systematic general practice: a new approach from the Tavistock clinic. Br J Gen Pract 1997;47: 453-6.(M Cardol, researcher1, P P Groenewegen, )