Qualitative and quantitative research methodsAuthor(s): Larry W. ChambersSource: Canadian Journal of Public Health / Revue Canadienne de Sante'e Publique, Vol. 88, No.1 (JANUARY / FEBRUARY 1997), pp. 9-10Published by: Canadian Public Health AssociationStable URL: http://www.jstor.org/stable/41992656 .

Accessed: 16/06/2014 23:31

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Canadian Public Health Association is collaborating with JSTOR to digitize, preserve and extend access toCanadian Journal of Public Health / Revue Canadienne de Sante'e Publique.

http://www.jstor.org

This content downloaded from 188.72.126.25 on Mon, 16 Jun 2014 23:31:36 PMAll use subject to JSTOR Terms and Conditions

LETTERS TO THE EDITOR

CORRESPONDANCE

The emerging interna-

tional dominance of

chronic diseases Dear Editor ,

The editorial by Ronald St. John in the most recent issue of the Journal (Volume 87, No. 6, November/December 1996), entitled "Emerging infectious disease: Repeat of an old challenge", alerts readers to this important national and international public health concern, but his assertion that "infectious diseases are still the leading cause of death worldwide" is simply not true.

In fact, chronic non-communicable dis- eases (CNCDs) have superseded communi- cable, maternal and perinatal conditions as the leading causes of death in all regions of the world except Sub-Saharan Africa (SSA) and the Middle East. CNCDs now domi- nate in all ages over 14 years, except in SSA.

For example, in Latin America and the Caribbean (LAC), the ratio of deaths from chronic to infectious and parasitic diseases was 1.5 in 1985, projected at 3.4 in the year 2000, rising to 6.7 in 2015. In LAC, chronic diseases now account for 57.9% of mortali- ty, with injuries an additional 9.8% (total 67.7%). These conditions also dominate estimates of years lost to disability, account- ing for 54.2% of the impact, plus injuries 17.7% (total 71.9%). The majority occurs in persons younger than 45 years of age (men 62.7%, women 65%). Approximately 30% of this disability impact is among per- sons younger than 1 5 years of age.

Contrary to popular myth, therefore, the bulk of disease in the world today is due to chronic non-communicable diseases, more of which are found in developing as opposed to developed countries. These conditions (e.g., heart disease, cancer, dia- betes, brain injuries) are not concentrated among the wealthy, but are actually more associated with poverty in all countries. Most of the burden is premature, and much is preventable. International recogni- tion, at a level accorded to the infectious diseases despite their overall decline in most regions, would do a lot to promote

aid priorities that fully reflect the facts of the international health situation.

Franklin White, MD , CM, MSc, FRCPC Program Coordinator N on-Communicable Diseases Pan American Health Organization

REFERENCES 1. Murray CJL, Lopez AD (Eds.). Global

Comparative Assessments in the Health Sector. Geneva: World Health Organization, 1994.

2. Pan American Health Organization. Health Conditions in the Americas , 1994 Edition. Volume 1. Washington, DC: PAHO, 1994 (Scientific Publication No. 549).

Response from author My assertion that "infectious diseases are

the leading cause of death worldwide", which is challenged by Dr. Frank White in the above letter, is taken directly from the World Health Organization's Annual Health Report for 1996.

Ron St. John, MD, MPH Director, Office of Special Health Initiatives Health Canada

Qualitative and quanti- tative research methods

Dear Editor, I am making a plea that the CJPH con-

tinue to publish a "range of research". Qualitative field research is relevant in a wide variety of public health activities where the distinguishing characteristic of the phenomena being studied involves a personal closeness and understanding between the public health practitioner and the client/citizen as an essential part of their interaction.

Criteria for assessing the validity and reliability of such field studies require that the investigators have: • asked subjects if the observations about

them are credible • had prolonged engagement by the

observers to minimize distortions caused by their presence

• conducted triangulations, which involved pitting against each other dif- ferent data and theoretical interpreta- tions, so as to provide crosschecks of observations and interpretations (Polgar S, Thomas SH. Introduction to Research in the Health Sciences. New York: Churchill Livingstone, 1995). My colleagues involved with producing

systematic reviews on public health topics are working with the Cochrane Health Promotion Field within the Cochrane Collaboration to develop criteria for research in this area. A number of issues are being pursued including development of criteria for assessing the impact when the health promotion intervention has multiple components. This work, like our paper on location of randomized con- trolled trials (RCTs) in the CJPH? makes the plea that RCTs should be used in pub- lic health whenever possible.

Advocating for RCTs in public research, however, is not inconsistent with the fact that there are a number of reasons why research designs other than RCTs may be employed: • Many variables are not amenable to

experimental manipulation. For exam- ple, if the research question is concerned with gender differences in responses to a new vaccine, then gender cannot be manipulated by the researcher. Similarly, if the researcher is interested in age dif- ferences, the ages of the participants can- not be altered by him or her. Many such variables cannot be manipulated and hence cannot be incorporated in this way in an RCT.

• Often it is ethically inappropriate to investigate research questions using an RCT. For example, if a researcher wished to perform a study on the effects of smoking upon health, to do this in an RCT would require the experiment to randomly allocate participants to the smoking or non-smoking group, i.e., force people to smoke and others not to smoke. In RCTs using a non- intervention control group, valuable and effective intervention might be withheld from participants.

• RCTs are best used to study simple causal relationships between variables. Yet many human diseases and illnesses

JANUARY - FEBRUARY 1997 CANADIAN JOURNAL OF PUBLIC HEALTH 9

This content downloaded from 188.72.126.25 on Mon, 16 Jun 2014 23:31:36 PMAll use subject to JSTOR Terms and Conditions

LETTERS TO THE EDITOR

are not determined by a single cause but rather by a number of causes interacting in a complex fashion. For instance, heart disease appears to be caused by factors such as low socioeconomic status, smok- ing, excessive stress, inappropriate diet and genetic factors. To identify such possible causal (or risk) factors, we need to study systems as they function in nature. That is, we should investigate clients/citizens in their natural setting, even with the difficulties this entails. Grounded theory, developed by Glaser

and Strauss ( The Discovery of Grounded Theory: Strategies for Qualitative Research. New York: Aldine, 1967), is a key approach to theoretical qualitative research. They advocate two methods for the development of grounded theory: the constant compara- tive method in which the researcher codes and analyzes data to develop concepts, and the theoretical sampling method in which cases are selected purposively to refine the 'theory' previously developed. This research is similar to qualitative field studies in that it shares common analysis tools such as coding and thematic analysis.

Larry W. Chambers , BA(Hons), MSc, PhDy FACE Epidemiology Consultant , Hamilton- Wentworth Regional Public Health Department; and Professor, Department of Clinical Epidemiology and Biostatistics McMaster University

REFERENCE 1. Chambers LW, Gold M, Krueger PD, et al.

Randomized trials of public health interventions reported in the Canadian Journal of Public Health: 1966 to 1996. Can J Public Health 1 996;87(6):4 1 1-12.

Rickets prevention

Dear Editor, Re: Binet A, Kooh SW. Persistence of vit-

amin D-deficiency rickets in Toronto in the 1990s. Can J Public Health 1996;87(4):227-30.

In response to my letter ( Can J Public Health 1996;87(6):375), Binet and Kooh agree that children born to mothers with poor vitamin D status are at greater risk for defi-

ciency. They do not agree that supplementa- tion is unnecessary if maternal stores are ade- quate. Obviously, just what would constitute adequate is debatable. However, the real problem lies with the evidence they provide to support their belief that "exclusively breast- fed, unsupplemented infants are at risk of developing vitamin D deficiency, even if maternal vitamin D nutrition is adequate."

To support their view, Binet and Kooh cite two articles reporting on research from the early 1980s.1,2 In both studies, there is no way to determine the vitamin D status of the mothers of children who developed low vitamin D levels. Greer et al. never measured the vitamin D status of the mothers (see note 1). A later randomized trial by Greer and Marshall3 concluded that "[u]nsupple- mented, human milk-fed infants had no evi- dence of vitamin D deficiency during the first six months of life." (see note 2)

Binet and Kooh also state that a study by Ala-Houhaha1 "showed that many infants born in winter to mothers with normal 25- OHD levels at delivery and who continued to take vitamin D supplements developed low serum 25-OHD concentrations by eight weeks of age." However, there is no way to know if the mothers (of the children who developed low levels) had normal lev- els at birth (see note 3). In an earlier study, Ala-Houhaha et al.4 found, as have other studies,5'6 that the 25-OHD concentrations in cord-blood, and in infants, is closely cor- related with maternal levels. It seems likely that the infants who went on to develop low levels were those born to women with low levels, but the information needed to determine this is lacking, just as the infor- mation needed to know if the mother had normal levels is also lacking.

Ala-Houhaha1 pointed out that the sum- mer infants stayed with in the normal range because the levels at delivery were high. This indicates the importance of levels at birth. Vitamin supplementation of the infant is the appropriate treatment for vitamin D defi- ciency and rickets. Where there is reason to suspect an infant is at risk (i.e., siblings with a history of rickets, evidence of other malnu- trition, low calcium and vitamin D intake of the mother), prophylactic use of vitamin D is an effective form of secondary prevention. However, primary prevention should focus on the health of the mother.

Pamela J. Smith PhD Student Dept. of Community Health Sciences Faculty of Medicine The University of Manitoba

REFERENCES 1. Ala-Houhaha M. 25-hydroxyvitamin D levels

during breast-feeding with or without maternal or infantile supplementation of vitamin D. J Pediatr Gastro Nutr 1985;4:220-26.

2. Greer FR, Searcy JE, Levin RS, et al. Bone min- eral content and serum 25 hydroxyvitamin D concentration in breastfed infants without sup- plemented vitamin D: one-year follow-up. J Pediatr 1982;100:919-22.

3. Greer FR, Marshall S. Bone mineral content, serum vitamin D metabolite concentrations, and ultraviolet light exposure in infants fed human milk with and without vitamin D2 supplements. J Pediatr 1 989; 1 14:204-12.

4. Ala-Houhaha M, Koskinen T, Terho A, et al. Maternal compared with infant vitamin D supple- mentation. Arch Disease Child 1986;6l:l 159-63.

5. Paunier L, Lacourt G, Pilloud P, et al. 25-hydrox- yvitamin D and calcium levels in maternal, cord and infant serum in relation to maternal vitamin D intake. Helv paediat Acta 1978;33:95-103.

6. Markestad T. Effect of season and vitamin D supplementation on plasma concentrations of 25- hydroxyvitamin D in Norwegian infants. Acta Paediatr Scand 1 983;72:8 1 7-2 1 .

7. Greer FR, Searcy JE, Levin RS, et al. Bone min- eral content and serum 25-hydroxyvitamin D concentration in breast-fed infants with and without supplemental vitamin D. J Pediatr 1981;98:696-701.

8. Clements MR, Johnson L, Fraser DR. A new mechanism for induced vitamin D deficiency in calcium deprivation. Nature 1987;325:62-65.

NOTES 1. Dietary recall studies were conducted post-

natally.7 Vitamin D status is determined by factors other than sun exposure and vitamin D intake. Vitamin D is involved in calcium homeostasis, and research since the publication of the cited papers has found that habitual low calcium con- sumption/absorption results in increased meta- bolic inactivation of vitamin D.8 Greer et al. reported that the mothers averaged less than the recommended intake of calcium.3

2. They found that, despite similar sun exposure to UVB rays and lower overall 25-OHD levels, 25- OHD3 levels were higher in the unsupplemented human milk-fed group. They proposed that the production of vitamin D3 was inhibited by high levels of 25-OHD2 in supplemented infants.3 This study included vitamin D analysis of cord blood at birth, which has been shown to correlate with maternal vitamin D status. All infants had con- centrations in the "adult normal range" at birth.

3. The three treatment groups all appear from the graph to have average levels in the normal range, but 12 of 47 mothers in the winter groups had levels below 5 ng/mL. Ten of eighteen unsupple- mented children born in winter developed low levels. Whether any of the infants had low levels at birth is not reported, only that they were simi- lar for all three groups in the winter and signifi- cantly lower than in the summer groups.3

10 REVUE CANADIENNE DE SANTÉ PUBLIQUE VOLUME 88, NO. 1

This content downloaded from 188.72.126.25 on Mon, 16 Jun 2014 23:31:36 PMAll use subject to JSTOR Terms and Conditions