Family Practice Advance Access originally published online on July 15, 2005
Family Practice 2005 22(6):658-662; doi:10.1093/fampra/cmi066
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Predicting parents' decisions on MMR immunisation: a mixed method investigation
a School of Nursing, University of Manchester and b School of Social Sciences, Media & Communication, Queen Margaret University College, Edinburgh, UK.
Correspondence to Dr Chris McVittie, School of Social Sciences, Media & Communication, Queen Margaret University College, Clerwood Terrace, Edinburgh EH12 8TS; Email: cmcvittie{at}qmuc.ac.uk
Received 21 October 2004; Accepted 20 June 2005.
Gellatly J, McVittie C and Tiliopoulos N. Predicting parents' decisions on MMR immunisation: a mixed method investigation. Family Practice 2005; 22: 658–662.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Background. Increasing uptake rates for MMR vaccination requires an understanding of factors leading parents to decide for and against vaccination, particularly in the light of recent developments.
Objective. We investigated factors relevant to immunising and non-immunising parents and the extent to which these factors predicted their decisions.
Methods. The study was conducted in Edinburgh, UK. A two-stage mixed method design was used. Delphi technique elicited parents' views of factors relevant to MMR immunisation. Twenty-six factors identified as relevant were incorporated into a final questionnaire. Using cluster sampling, the final questionnaire was distributed to parents recruited through a group of local nurseries. 110 parents participated: eighty (72.7%) had had their child MMR vaccinated, thirty (27.3%) had refused the vaccine. The factors in the final questionnaire were analysed against vaccination status using a direct binary logistic regression model.
Results. Four factors significantly predicted vaccination status, (prediction toward ‘yes vaccination’). These were the influence of current research (OR = 0.18, 95% CI = 0.07–0.51), the helpfulness of leaflets and information packs (OR = 3.27, 95% CI = 1.38–7.75), the importance of eradication of rubella (OR = 2.42, 95% CI = 1.01–5.78), and the importance attached to the risk of adverse reactions (OR = 0.65, 95% CI = 0.48–0.87).
Conclusions. Differences between immunising and non-immunising parents lie in the importance attached to four relevant factors. Excluding risk of adverse reactions, these factors have not been previously identified as salient and require to be explored further. Health advice to parents should highlight the identified importance attached to eradicating rubella and explicitly reflect research findings.
Keywords. Attitudes, immunization, MMR vaccine, parents.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Recent developments make this a particularly apposite time to consider parental attitudes towards MMR vaccination of their children. Parental concerns over safety of the vaccine have been clearly visible since the initial suggestion by Wakefield and colleagues1 of a possible link between children's receipt of the vaccine, gastrointestinal disease and autistic-spectrum disorders. Publication of this suggested link was followed by a decline in uptake rates of the combined MMR vaccine across the UK, in some parts to as low as 61.4%.2 Notwithstanding repeated criticisms of the Wakefield study on a number of grounds, for example that the findings relied solely on temporal coincidence,3 that an incorrect reference range was used,4 and that the claims are not supported by epidemiological evidence5–10 or findings from matched control studies,11 uptake rates remain at close to their lowest ever levels. In England, for example, immunisation rates have fallen from a peak of 92% in 1995–96 to 80% in 2003–04, with a current uptake rate in south-east London of only 62%.12 It remains to be seen if criticism of the 1998 study's findings as being of questionable validity due to ‘potential conflict of interest’13 and the retraction by many of the authors of the 1998 study of the interpretation placed upon their findings14 will have any greater effect in restoring public confidence in the vaccine and increasing immunisation rates.
Restoring confidence effectively and achieving an increase in immunisation rates requires an understanding of the factors that have led many parents to decide against MMR vaccination of their children. Previous studies have identified to some extent factors relevant to parents for and against vaccination. For example, a comparison of immunising and non-immunising parents found differences in assessments of risks and benefits of vaccination, confidence in health professionals' advice and the importance attached to individual choice in the matter of vaccination.15 Studies of non-immunising parents have identified concerns over safety, the likelihood of contracting disease and a lack of belief in health professionals' advice as potential factors.16 By contrast, a survey of vaccine coverage across all health authorities in England and of attitudes of mothers within 132 enumeration districts concluded that most mothers continued to seek and be guided by advice from health professionals.17 Although such studies point to differences between immunising and non-immunising parents, it remains unclear what weightings are attached to these differences and how they are reflected in the immunisation decisions of each group. What does seem clear is that parents' decisions are not made on the basis of a risk/benefit analysis, either objective18,19 or subjective20 but rather reflect an array of personal attitudes, beliefs and perceptions.15,21 In this, parental behaviour is consistent with that found in relation to vaccination generally22 and other health behaviours involving perceptions of risks.23
We report here the results of a study of the factors that predict parents' decisions for and against MMR vaccination of their children, in the light of renewed criticism of the Wakefield et al. study.1 In this study we aimed (a) using a ‘bottom-up’ approach, to identify the factors relevant to both immunising and non-immunising parents, and (b) to investigate the extent to which the identified factors predicted immunisation or non-immunisation.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Setting
The study was conducted in Edinburgh, UK between December 2003 and May 2004. An institutional ethics committee granted ethical approval for the research.
Participants
A total of 415 children day-care nurseries are registered in Edinburgh. Of these, many offer access only to restricted groups of children, e.g. organizational nurseries that offer access only to children of employees. One group comprising five nurseries was selected, because the nurseries in this group did not restrict access to particular groups and the nurseries were located in postcode areas with a range of socio-economic characteristics that reflected those found within the local population24 and spanned the classes of The National Statistics Socio-economic Classification 2001.25 Subsequently, a cluster sampling approach was used, where all parents (n = 182) whose children attended these nurseries were invited to participate.
Design
We used a two-stage mixed design, comprising Delphi technique followed by attitude questionnaires.
Stage 1. In the first stage, factors relevant to immunisation decisions were elicited from the participants using two rounds of Delphi technique.26,27 The Delphi technique facilitates the aggregation of individual views on a topic and enables the researcher ‘to explore or expose underlying assumptions or information leading to differing judgments’.28 The aim of using Delphi technique was to identify elements relevant for the participants themselves rather than focusing on elements pre-determined by the researchers as being important. In the first Delphi round we used a questionnaire containing 6 open-ended items that reflected factors identified as relevant in previous research, namely: perceived risks and benefits of the MMR vaccine; importance of the combined vaccine; influences on the decision whether or not to immunise, and advice and information found to be helpful and unhelpful in making the decision. A further item invited additional comments. Participants were also asked if their children had (immunising parents) or had not received the MMR vaccine (non-immunising parents).
The questionnaire was completed by a sub-sample of 15 randomly selected individuals from the parents that had agreed to participate, comprising 8 immunising (53.3%) and seven non-immunising (46.7%) parents. Content analysis of responses, conducted independently by all authors and using inductive category construction,29 showed high consensus of views among both immunising and non-immunising parents on the factors relevant to each item. From the analysis, a second questionnaire was prepared. Participants were advised that the items included in this questionnaire had been identified from responses to the first questionnaire as primary relevant factors. The sample comprised the 15 participants responding in the first Delphi round. Participants were required to rank the factors included (1 = most important, 10 = least important). Again participants were invited to express additional comments, whether agreeing or disagreeing with initial analysis.
Structuring content analysis29 of second-round responses was again conducted independently by all authors. Analytic differences and apparent negative cases were discussed and the analysis refined as necessary. Final analysis showed a high level of agreement among both immunising and non-immunising parents in rankings of factors relevant to each item. The dimensions of structuration produced by this analysis were used to generate items for inclusion in the instrument used in the second stage of the study.
Stage 2. The final questionnaire comprised 26 questions, which accessed the degree of importance or helpfulness the participants assigned to the factors identified earlier. All responses were registered on a 5-point Likert scale. These questions formed four conceptual groups, which addressed:
- The importance of the perceived risk of someone immunised with the MMR vaccine to develop autism, bowel disease, Crohn's disease, other allergic reactions, or to overload their immune system (five questions; 1 = very unimportant to 5 = very important).
- The importance of the following benefits of the MMR vaccine: eradications of measles, mumps, and rubella, separate protection from each of the above, and protection from all the above diseases in one dose (five questions; 1 = very unimportant to 5 = very important).
- The importance of the following factors on the decision to vaccinate: research, opinion of GPs and health visitors, lack of information, nature of long-term effects, and safety of the vaccine (six questions; 1 = very unimportant to 5 = very important).
- The helpfulness of the following set of factors toward the decision to vaccinate: conflicting opinions, experiences of friends/family, government health advice, government pressure, GPs' and health visitors' advice, lack of clear evidence, lack of information on single vaccines, information packs, and media (ten questions; 1 = very unhelpful to 5 = very helpful).
- The importance of the following benefits of the MMR vaccine: eradications of measles, mumps, and rubella, separate protection from each of the above, and protection from all the above diseases in one dose (five questions; 1 = very unimportant to 5 = very important).
Statistical Analysis
SPSS (v.11.5) was used to carry out statistical analysis. The relationship between vaccination status (dependent dichotomous variable), and the 26 questions of the four conceptual groups was assessed at the bivariate level through a series of point-biserian correlations.30 Only variables that had statistically significant correlation coefficients at the 0.05 level (two-tails) were considered as potential predictors of vaccination status. Moreover, Pearson's product moment correlations were used to assess the degree of multicollinearity between those predictors. Variables with associations that had correlation coefficients greater than r = ±0.70 were assumed to possess significant levels of collinearity. Such variables were grouped into conceptually meaningful factors and each participant was given a mean response on each factor. These factors and the rest of the statistically significant predictors were subsequently used in a direct binary logistic regression model against vaccination status.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Of the parents contacted, 110 (60.4%) agreed to participate. The median age of the participants' youngest child was 21 months (range one month to 59 months), with 49 (44.5%) of them being female. All parents had at least one child of vaccination age. Eighty participants (72.7%) had had their child vaccinated with the MMR vaccine, while the rest had refused the vaccine. Five participants who returned incomplete responses were excluded from the final analysis.
Seventeen variables had a statistically significant relationship with vaccination status (P < 0.05, two-tails). Those were: the importance of the perceived risk of someone immunised with the MMR vaccine to develop (1) autism, (2) bowel disease, (3) Crohn's disease, (4) other allergic reactions, and (5) to overload the immune system; importance of protection from (6) rubella, and (7) measles, mumps, and rubella in one dose; (8) influence of current research, (9) GPs' opinion, (10) health visitors' opinion, (11) lack of information on single vaccines, and (12) nature of long-term effects; helpfulness of (13) government pressure, (14) government health advice, (15) GPs' advice, (16) health visitors' advice, and (17) information packs.
In order to minimise multicollinearity problems variables (1), (2), (3) and (5) were grouped into a single factor, which was termed ‘risk from adverse reactions’, GP's (9) and health visitors' (10) opinion were grouped under a ‘health professionals' opinion’ factor, while GPs' (15) and health visitors' (16) advice formed a ‘health professionals' advice’ factor. The above grouping reduced the number of variables to twelve.
A direct binary logistic regression was performed on vaccination status as outcome and the above twelve variables as predictors. The overall model, presented in Table 1, was statistically reliable, 
2 (12, n = 105) = 77.20, P < 0.0005, with an acceptable goodness-of-fit, Holmer & Leneshow chi-square (8, n = 105) = 7.39, P = 0.49. The variance in vaccination status accounted for was 75% (Nagelkerke R2), while the model correctly classified 80% of the ‘no vaccination’ status and 93.30% of the ‘yes vaccination’ status, with an overall success rate of 89.20%.
|
In the model, four predictors had statistically significant ORs. The influence of current research was the strongest predictor, indicating that parents who viewed research findings as more important were approximately five and a half times as likely not to vaccinate their children than those who did not (OR = 0.18, 95% CI = 0.07–0.51, prediction toward ‘yes vaccination’ status). Parents who found useful the information contained in leaflets and packs were more than three times as likely to vaccinate their children than those who did not (OR = 3.27, 95% CI = 1.38–7.75). When the eradication of rubella was perceived as more important, it increased the likelihood of vaccination by 2.4 times (OR = 2.42, 95% CI = 1.01–5.78). Finally, parents who viewed the risk of adverse reactions as more important were approximately one and a half times as likely not to vaccinate their children than those who did not (OR = 0.65, 95% CI = 0.48–0.87, prediction toward ‘yes vaccination’ status).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Both immunising and non-immunising parents initially identified similar factors as being relevant to decisions on MMR vaccination. The differences between these groups lie in the importance attached to four factors in making their decisions. Of these factors, two predict decisions in favour of vaccination. Parents who attach importance to the eradication of rubella and to information packs on vaccination are 2.4 times and more than three times respectively more likely to have their children vaccinated than parents who do not. Two factors, namely the risk of adverse reactions and the importance of current research findings, predict parental decisions against vaccination by odds of approximately one and a half to one and five and a half to one, respectively. The reverse predictive directions of these two factors and of the two factors discussed above indicate that both immunising and non-immunising parents view the advice contained in information packs as being distinct from current research findings on this topic.
This study provides a more detailed understanding of differences between immunising and non-immunising parents than that found in single method studies. It identifies where the differences between these groups of parents lie and how these differences predict parental decisions for and against MMR vaccination of their children. Additionally it provides an up-to-date analysis of parental views on MMR vaccination following recent developments. The research however is confined to the study of immunisation decisions within one geographical location and needs to be extended to obtain a more general picture.
Previous studies have found differences between immunising and non-immunising parents in the importance that they attach to health professionals' advice.15 Here we found such differences to be significant but not predictive of vaccination decisions. The predictive role identified here of another source of health advice, namely information packs and leaflets, has not previously been identified.
Similarly, the weight given to the eradication of rubella has not previously been identified as a salient factor. Previous work relating to decreasing MMR immunisation rates has emphasised measles as the primary cause for concern.31–33 In this, the views of professionals appear to differ from those of immunising parents.
The perceived risk of adverse effects as predictive of non-immunisation decisions supports findings of differences between immunising and non-immunising parents in their perceptions of risks15 and of safety concerns as a major factor for non-immunising parents.16 Perception of research findings as a discrete factor in non-immunisation decisions has not been identified in previous research. The predictive direction of this factor is however consistent with the finding that parents regard health advice on this issue as uninformed by relevant research findings.34
Further research is needed to examine in greater detail the factors identified here. It appears unlikely, for example, that the majority of parents are familiar first-hand with much of the research on this topic. Instead the perceptions of non-immunising parents might well reflect selective media coverage of the topic, which is often presented in somewhat partial ways.35 Research consequently is required to explore in greater detail the formation of parental perceptions both of research findings and of the risk of adverse effects.
In terms of practice, these findings suggest that the advice currently provided by information packs is found useful by immunising parents and influences their decisions. The importance attached by immunising parents to the eradication of rubella however offers some scope for development of practice. Increased emphasis on this factor in future advice potentially might increase the number of parental decisions in favour of vaccination.
Of greatest concern though is the finding that both immunising and non-immunising parents perceive the health advice they receive on this issue to be entirely distinct from relevant research findings. More needs to be done to demonstrate that current advice is informed by and reflects the state of accumulated research knowledge of the topic. The currently perceived disjunction challenges both practitioners to demonstrate the application of research findings, and researchers to produce findings that can be directly disseminated to parents. Only then are these different sources of information likely to be viewed by parents as providing consistent and useful advice, on which they can base their decisions.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Funding: no external funding was received.
Ethical approval: an institutional ethics committee granted ethical approval for the research.
Conflicts of interest: none.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
1 Wakefield AJ, Murch SH, Anthony A et al. Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet 1998; 351: 637–641.[CrossRef][ISI][Medline]
2 Kidd IM, Booth CJ, Rigden SPA, Tong CYW, MacMahon EME. Measles-associated encephalitis in children with renal transplants: a predictable effect of waning herd immunity. Lancet 2003; 362: 832.[Medline]
3 Farrington CP, Miller E, Taylor B. MMR and autism: further evidence against a causal association. Vaccine 2001; 19: 3632–3635.[CrossRef][ISI][Medline]
4 Chen RT, DeStefano F. Vaccine adverse events: causal or coincidental? Lancet 1998; 351: 611–612.[CrossRef][ISI][Medline]
5 Medical Research Council. Report of the strategy development group subgroup on research into inflammatory bowel disorders and autism. London: The Council; 2000.
6 Medical Research Council. Review of autism research: epidemiology and causes. 2001. Available from: URL: http://www.mrc.ac.uk/pdf-autism-report.pdf
7 Taylor B, Miller E, Farrington CP, Petropoulous MC, Favot-Mayand LJ, Waight PA. Autism and measles, mumps and rubella vaccine: no epidemiological evidence for a causal association. Lancet 1999; 353: 2026–2029.[CrossRef][ISI][Medline]
8 Taylor B, Miller E, Lingham R, Andrews N, Simmons A, Stowe J. Measles, mumps, and rubella vaccination and bowel problems or developmental regression in children with autism: population study. Br Med J 2002; 324: 393–396.
9 Chen W, Landau S, Sham P, Fombonne E. No evidence for links between autism, MMR and measles virus. Psychological Med 2004; 34: 543–553.
10 Honda H, Shimizu Y, Rutter M. Reported in Coghlan A. Autism rises despite MMR ban in Japan. New Scientist; 3 March 2005.
11 Smeeth L, Cook C, Fombonne E, Heavey L, Rodrigues LC, Smith PG, Hall AJ. MMR vaccination and pervasive developmental disorders: a case-control study. Lancet 2004; 364: 963–969.[CrossRef][ISI][Medline]
12 Department of Health. Statistical Bulletin: NHS Immunisation Statistics 2003–04. London: Department of Health; 2004.
13 Horton R. A statement by the editors of The Lancet. Lancet 2004; 363: 820.[CrossRef][ISI][Medline]
14 Murch SH, Anthony A, Casson DH et al. Retraction of an interpretation. Lancet 2004; 363: 750.[CrossRef][ISI][Medline]
15 Evans M, Stoddart H, Condon L, Freeman E, Grizzell M, Mullen R. Parents' perspectives on the MMR immunisation: a focus group study. Br J Gen Pract 2001; 51: 904–910.[ISI][Medline]
16 Johnson A, Joynes E. MMR vaccination uptake in a rural setting. Br J Gen Pract 2001; 51: 408.[Medline]
17 Ramsay ME, Yarwood J, Lewis D, Campbell H, White JM. Parental confidence in measles, mumps, and rubella vaccine: evidence from vaccine coverage and attitudinal studies. Br J of Gen Pract 2002; 52: 912–916.
18 Murch SH. Separating inflammation from speculation in autism. Lancet 2003; 362: 1498–1499.[ISI][Medline]
19 McMurray R, Cheater FM, Weighall A, Nelson C, Schweiger M, Mukherjee S. Managing controversy through consultation: a qualitative study of communication and trust around MMR vaccination decisions. Br J Gen Pract 2004; 54: 520–525.[Medline]
20 Bellaby P. Communication and miscommunication of risk: understanding UK parents' attitudes to combined MMR vaccination. Br Med J 2003; 327(7417): 725–728.
21 Petousis-Harris H, Turner N, Kerse N. New Zealand mothers' knowledge of and attitudes towards immunisation. NZFP 2002; 29(Pt 4): 240–246.
22 Leask J-A, Chapman S, Hawe P. Concerns about immunisation. Facts are not enough. Br Med J 2000; 321: 109.
23 Welton M, Hepworth J, Collins N, Ford D, Knott C, Meredith S, Walgrove A, Wilkes H, Vickers M. on behalf of the WISDOM team. Decision-making about hormone replacement therapy (HRT) by women in England and Scotland. Climacteric 2004; 7: 41–49.[CrossRef][Medline]
24 Danesh J, Gault S, Semmence J, Appleby P, Peto R. Postcodes as useful markers of social class: population based study in 26 000 British households. Br Med J 1999; 318: 843–845.
25 Office for National Statistics. The National Statistics Socio-economic Classification. London: Office for National Statistics; 2001.
26 Dalkey NC. The Delphi method: an experimental study of group opinion. Memorandum RM 5888-PR. Santa Monica: RAND Corp.; 1969.
27 Dalkey NC. An experimental study of group opinion: the Delphi method. Futures 1969; 9: 408–426.
28 Turoff M. The design of a policy Delphi. Technological Forecasting and Social Change 1970; 2: 149–171.
29 Sarantakos S. Social research. 3rd edn. Basingstoke: Palgrave; 2005.
30 Stuart A, Ord K. Kendall's Advanced Theory of Statistics Vol. 2. 5th edn. London: Arnold; 1991.
31 Kidd IM, Booth CJ, Rigden SPA, Tong CYW, MacMahon EME. Measles-associated encephalitis in children with renal transplants: a predictable effect of waning herd immunity. Lancet 2003; 362: 832.[Medline]
32 Jansen VA, Stollenwerk N, Jensen HJ, Ramsay ME, Edmunds WJ, Rhodes CJ. Measles outbreaks in a population with declining vaccine uptake. Science 2003; 301: 804.
33 Fitzpatrick M. MMR: risk, choice, chance. Br Med Bull 2004; 69: 143–153.
34 Smailbegovic MS, Laing GJ, Bedford H. Why do parents decide against immunization? The effect of health beliefs and health professionals. Clinical Care Health and Development 2003; 29(4): 303–311.
35 see e.g. Elliman D, Bedford H. Hear the Silence. Br Med J 2003; 327: 1411.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Delaney Family Practice and the new era of electronic publishing Fam. Pract., December 1, 2005; 22(6): 581 - 582. [Full Text] [PDF]
|
|
E-letters:
Read all E-letters
- Good enough reasons?
- Alan Challoner
- Family Practice, 28 Nov 2005 [Full text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||