Critical assessment of methodology and quality

Critical evaluations of the studies included in this review were done according to the guidelines of Loney & Stratford (1999) and Loney et al (1998). Among eight parameters of the guidelines, the “sample size” parameter was adjusted for this review, as the original guidelines referred to dementia which is a relatively more rare disease than diabetes. Using a conservative sample size estimate of proportion for this review of dementia, prevalence of use of homeopathy in diabetic patients, (assumptions based on Mehrotra et al, 2004) the adequate sample size has been set at ?450. Detail scoring for methodology of the studies is shown in table 3, and discussed below.

Table 3 Quality assessment

A survey (observational study) is the appropriate study design to determine the prevalence of particular health problems or use of any therapy. If the whole population of interest is not surveyed, then the best sampling technique is random (probability) sampling of persons from a defined subset of the population. Stratification (sampling purposely from subgroups) may be required to appropriately represent subgroups (O’Rourke, 2005, Sim & Wright 2000). All the studies included in this review are survey and the sampling procedure is appropriate (Table 3, parameter 1). For larger surveys, cluster sampling is sometimes used as employed by Dannemann et al, (2008) one of the studies included in this review. In cluster sampling, groups of individuals are selected as the survey unit. Dannemann et al, surveyed four pediatric diabetes centerers in Germany, two from west Germany (Bonn and Stuttgart) and two from East Germany (Leipzig and Berlin).

Type of sampling frame from which subjects are selected is important (Hennekens & Buring 1987). Census data provide one of the few data sets from which one can draw a sample that is thought to have minimal bias, since certain groups of persons are thought not to be excluded as they might be in an electoral list or telephone list (Sica, 2003). Only one study (Yeh et al, 2002) has used census data for sampling among the studies included in this review (Table 3, parameter 2). The rest of the three studies were conducted in a diabetic clinic, limiting their generalisability over a greater population.

A large sample size produces narrow confidence limits, which is undoubtedly important if the prevalence of a given condition is low. Small sample sizes produce large confidence intervals, making the findings less precise. It is critical to be as confident as possible that any changes in health care policy are based on results that did not occur by chance due to probability sampling inadequacy. (Slavin, 1995) Sample size required to estimate a proportion with a specified degree of precision (for example 95% confidential intervals) can be calculated (Katchigan, 1986: pp 158-9). Using a conservative sample size estimate of proportion for this review of dementia, prevalence of use of homeopathy in diabetic patient (assumptions based on Mehrotra et al, 2004) the adequate sample size has been taken as ?450. In only two studies adequate samples have been surveyed (Yeh Table 3 parameter 3).

It is important that published studies describe the measurement units well enough so that the outcome measures can be compared (Grimes & Schulz, 2002). Since health problems can be defined in many ways, the measurement of the problem must be the best possible one (Greenhalgh, 2006). In the prevalence of use of homeopathy in diabetic patients, surveys are based on interview and self reported prevalence was recorded. In cases with this type of self reported prevalence, recall bias is a potential problem that may distort the result of the study. Recall bias occurs when exposure information is differentially misclassified for subjects with and for those without the condition under examination (Rothman, 2002: pp 94-112.). Recall bias can be particularly problematic in studies where subjects are interviewed to collect information (Sica, 2003). In this review, all the included studies were scored 1 point for appropriate measurement (table 3 parameter 4). However, Dannemann et al, (2008) should get extra weight for this parameter as they mention recall bias as the limitation of the study, which indicates that the researchers were aware of this bias.

Considerable judgment by assessors is required to determine the presence of some health outcomes under scrutiny; thus it is best that trained assessors are independent and not aware (i.e. blinded) of the subjects’ clinical status and the purpose of the study. It is important that the subjects under assessment include those thought to be negatives as well as positive (Lijmer et al, 1999). In case of the studies under review, no studies reported anything about the blinding of the interviewer (table 3 parameter 5). It could introduce serious bias if the interviewers are aware of the study’s purpose prior to the study, as the interviewers may have an inclination for or against homeopathy.

The greater the numbers of selected subjects who are lost to follow-up, the less valid the estimates are. A response rate in population surveys of two thirds to three quarters has been suggested to be generalizable to the population samples (Marshall, 1987). In this review a response rate of 70% has been chosen as acceptable. Since a large number of dropouts, refusals or “not founds” among the subjects selected may jeopardize a study’s validity, the authors should describe the reasons for non-response and compare persons in the study with those not in the study as to their socio-demographic characteristics (Response bias – Sica, 2003). If the reasons for non-response seem unrelated to the outcome measured and the characteristics of those individuals not in the sample are comparable to those in the study, researchers may be able to justify a more modest response rate (Loney et al, 1998). Among the four studies included in this review, only one study reported adequate response rate (table 3 parameter 6), while other studies did not even described the refusers.

The seventh parameter of the quality assessment is the estimate of prevalence of use of homeopathy in diabetic patients given with confidence intervals (CI) and in detail by subgroup or not. The quantitative results from studies of prevalence are proportions or rates over a fixed period of time (Szklo & Nieto, 2000). The prevalence rates found in studies reviewed provide only estimates of the true prevalence of use of homeopathy in the larger population. Confidence intervals then indicate the level of confidence one can have in the estimates and their range (Oliveira et al, 2006). Since some subgroups are very small, 95% confidence intervals have been taken as standard. Among the four studies included in this review only one study (Mehrotra et al, 2004) did not mention CI nor describe the subgroup.

Certain diseases and health issues are known to vary in prevalence across different geographic regions and population sectors. The status of homeopathy also varies by country and region (ECH, 2007).  With some health problems, rates for women may differ from those for men. Moreover, socio-demographic variables, such as educational status, may vary between countries. Therefore, the study sample needs to be described in enough detail that other researchers can determine if it is comparable to the population of interest to them. Furthermore, the socio-demographic characteristics of the subjects must be reported in order to understand the applicability of the results. Similarly, providing a comparison of study participants with those who refused or were ineligible can help others determine for whom the study group is representative. All studies included in this review have described their subjects and refusal except one study (parameter 8).

Overall, most of the studies (3 out of 4) scored 4 while the highest score was 8 for eight methodological parameters. In that sense, most of the studies are of average quality. Only one study scored 6 (Yeh et al, 2002). Sampling method of all the studies was unbiased, but no studies measure the outcomes by unbiased assessors (no blinding).

Key findings

Discussions & Conclusion

In this review, the prevalence range of use of homeopathy among diabetic patients, based on the findings of four papers, was 0.7 % to 12.9% with the lowest prevalence found in the USA and the highest in India. The reported range is in accordance with the report of ASSOCHAM (The Associated Chambers of Commerce and Industry of India) “Homeopathy is an effective means of treating chronic ailments. These ailments include …… diabetes, and obesity.” “.. reasons for growing homeopathy market in India, saying that homeopathy, besides providing an effective means of treating chronic ailments is also available and easily accessible online to over 1 crore patients across the country.” (ASSOCHAM, 2007).

Level of education increases the probability of use of homeopathy for diabetes and goes against the traditional belief that people use CAM (including homeopathy) due to fewer side effects with this type of treatment. Geographic area is also a significant predictor for use of CAM (and homeopathy). A large scale population-based survey or cohort study is needed to find out why diabetic patients use homeopathy, and what their expectation of homeopathy are, for the better management of diabetic patients.

The interpretation of the findings of this review are subject to a series of limitations. First of all, as any other systematic literature review, data was not collected by the author, which implies an over-reliance on the veracity of the data published. Secondly the number of papers included in analysis were only four, which prevents the author from producing strong statements about the final results. No meta-analysis was attempted for which the results of pooled data are not available here. Also, some of the inclusion/exclusion criteria can be interpreted as biases, e.g. only published articles and only studies published  in English. In spite of all these limitations, the author believes that this review will encourage more population based studies and reviews, to find an overview of the prevalence of use of homeopathy and its pattern and correlates in diabetic patients.

References

ASSOCHAM – The Associated Chambers of Commerce and Industry of India (2007) Homeopathy Emerging With Big Bang, Likely To Be Rs. 26 Billion Industry: ASSOCHAM, Sunday, December 09, 2007 [Online] Available at: http://www.assocham.org/prels/shownews.php?id=1308 [Accessed on: 10/07/2009]

Begg CB, McNeil BJ. (1988) Assessment of radiologic tests: control of bias and other design considerations. Radiology; 167: pp565 – 569.

Bell R.A., Suerken C.K., Grzywacz J.G., Lang W., Quandt S.A. & Arcury T.A. (2006) Complementary and alternative medicine use among adults with diabetes in the United States. Alternative Therapies in Health and Medicine; 12(5): pp16-22.

BMJ (1996) Complementary medicine is booming worldwide. BMJ; 313: pp131-133.

Chang H Y, Wallis M, Tiralongo E. (2007) Use of complementary and alternative medicine among people living with diabetes: literature review. Journal of Advanced Nursing; 58(4): pp307-19.

Dannemann K, Hecker W, Haberland H, Herbst A, Galler A, Schäfer T, Brähler E, Kiess W, Kapellen TM. (2008) Use of complementary and alternative medicine in children with type 1 diabetes mellitus – prevalence, patterns of use, and costs. Pediatric Diabetes; 9(3 Pt 1): pp228-35

Dunning T (2003) Complementary therapies and diabetes. Complementary Therapies in Nursing and Midwifery; 9(2): pp74-80.

ECH – European Committee for Homeopathy (2007) The position of Homeopathy in Europe.  [Online] Available at: http://www.homeopathyeurope.org/pdf/positionhomeopathyEurope.pdf [Accessed on: 10/07/2009]

Egede L E (2004) Complementary and alternative medicine use with diabetes. Geriatric Times; 5(2): pp

Eisenberg D.M., Davis R.B., Ettner S.L., Appel S., Wilkey S., Van Rompay M. et al. (1998) Trends in alternative medicine use in the United States, 1990-1997: Results of a follow-up national survey. JAMA: The Journal of the American Medical Association; 280(18): pp1569-1575.

Emslie M., Campbell M. & Walker K. (1996) Family Medicine Complementary therapies in a local healthcare setting. Part 1: is there real public demand? Complementary Therapies in Medicine; 4(1): pp39-42.

Garrow D, Egede L E. (2006a) National patterns and correlates of complementary and alternative medicine use in adults with diabetes. Journal of Alternative and Complementary Medicine; 12(9): pp895-902

Garrow D, Egede L E. (2006b) Association between complementary and alternative medicine use, preventive care practices, and use of conventional medical services among adults with diabetes. Diabetes Care; 29: pp15-19

Greenhalgh T., (2006) How to read a paper: the basics of evidence-based medicine. BMJ Books & Blackwell Publisher, United Kingdom.

Grimes D. A., & Schulz K., F., (2002) Descriptive studies: what they can and cannot do. Lancet 359: pp. 145-49

Hasan S S, Ahmed S I, Bukhari N I and Wei Loon W C (2009) Use of complementary and alternative medicine among patients with chronic diseases at outpatient clinics. Complementary Therapies in Clinical Practice; ARTICLE IN PRES

Hennekens C. H., Buring J., E. (1987) Analysis of epidemiologic studies: evaluating the role of bias. In: Hennekens C., H., Buring J., E. and Mayrent S., L. eds. Epidemiology in medicine. Lippincott Williams & Wilkins. pp. 272-286.

Katchigan S (1986) Statistical analysis: an interdisciplinary introduction to univariate and multivariate methods. New York: Radius Press.

Kumara D., S. Bajajb, R. Mehrotra (2006) Knowledge, attitude and practice of complementary and alternative medicines for diabetes. Public Health; 120: pp705-711

Leese GP, Gill GV, Houghton GM (1997) Prevalence of Complementary medicine usage within a diabetes clinic. Practical Diabetes International; 140: pp207-8

Lew-Ting C.Y. (2003) Who uses non-biomedical, complement and alternative health care? Socio-demographic un-differentiation and the effects of health needs. Taiwan Journal Public Health; 22(3): pp155-166.

Lijmer JG, Mol BW, Heisterkamp S, et al (1999) Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 282: pp1061-1066.

Lim M.K., Sadarangani P., Chan H.L. & Heng J.Y. (2005) Complementary and alternative medicine use in multiracial Singapore. Complementary Therapies in Medicine; 13(1): pp16-24.

Lind BK, Lafferty WE, Grembowski DE, Diehr PK. (2006) Complementary and alternative provider use by insured patients with diabetes in Washington State. Journal of Alternative and Complementary Medicine; 12(1): pp71-7.

Loney P L, Stratford P W. (1999) The prevalence of low back pain in adults: a methodological review of the literature. Physical Therapy; 79(4): pp384-96.

Loney P L, Chambers LW, Bennett KJ, Roberts JG, Stratford PW (1998) Critical appraisal of the health research literature: prevalence or incidence of a health problem. Chronic Diseases in Canada; 19(4): pp170-6.

MacLennan A.H., Wilson D.H. & Taylor A.W. (2002) The escalating cost and prevalence of alternative medicine. Preventive Medicine; 35(2): pp166-173.

Marshall V (1987) Factors affecting response and completion rates in some Canadian studies. Canadian Journal of Aging; 1: pp385 – 401.

Mehrotra R, Bajaj S, Kumar D. (2004) Use of complementary and alternative medicine by patients with diabetes mellitus. The National Medical Journal of India; 17(5): pp243-5.

Oliveira G. J., D.D.S., M.Sc.; Cláudio R. Leles, D.D.S., M.Sc., Ph.D. (2006), Critical appraisal and positive outcome bias in case reports published in Brazilian dental journals, Journal of Dental Education, 70(8): pp. 869-74.

O’Rourke A., (2005) Critical appraisal. In Bowling A., and Ebrahim S., (eds) Handbook of health research method: investigation, measurement and analysis. Open University Press. Berkshire; England.

Pagán J A, Tanguma J. (2007) Health care affordability and complementary and alternative medicine utilization by adults with diabetes.  Diabetes Care; 30(8): pp2030-1

Pomposelli R, Piasere V, Andreoni C, Costini G, Tonini E, Spalluzzi A, Rossi D, Quarenghi C, Zanolin ME, Bellavite P. (2009) Observational study of homeopathic and conventional therapies in patients with diabetic polyneuropathy. Homeopathy; 98(1): pp17-25.

Rothman KJ. (2002) Biases in study design. In: Epidemiology: an introduction. New York, NY: Oxford University Press.

Schoenberg NE, Stoller EP, Kart CS, Perzynski A, Chapleski EE. (2004) Complementary and alternative medicine use among a multiethnic sample of older adults with diabetes. Journal of Alternative and Complementary Medicine; 10(6): pp1061-6

Sica G. T., (2006) Bias in research studies. Radiology 238(3): pp. 780-89

Siegel K, Narayan KV. (2008) The Unite for Diabetes campaign: Overcoming constraints to find a global policy solution. Globalization and Health; 4: p3

Sim J., Wright C., (2000) Research in health care: concepts, designs and methods. Stanley Thomes (Publishers) Ltd. Cheltenham, United Kingdom.

Slavin R. E., (1995) Best evidence synthesis: an intelligent alternative to meta-analysis. Journal of clinical epidemiology. 48(1): pp. 9-18

Szklo M., Nieto, F., J., (2000) Identifying noncausal associations: confounding. In: Epidemiology: beyond the basics, 2^nd ed. Jones & Bartlett Publishers, Sudbury.

Tindle H.A., Davis R.B., Phillips R.S. & Eisenberg D.M. (2005) Trends in use of complementary and alternative medicine by US adults: 1997-2002. Alternative Therapies in Health and Medicine; 11(1): pp42-49.

WHO -IDF (2004), Diabetes Action Now, World Health Organization, [Online]. Available at: http://www.who.int/diabetes/actionnow/en/DANbooklet.pdf [Accessed on: 10/11/2008]

Yeh GY, Eisenberg DM, Davis RB, Phillips RS. (2002) Use of complementary and alternative medicine among persons with diabetes mellitus: results of a national survey. American Journal of Public Health; 92(10): pp1648-52.