SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Objective  To examine interpretations of study information by women participating in ORACLE, a trial of antibiotics in preterm labour.

Design  Questionnaire survey sent to women recruited to the ORACLE trial.

Setting  United Kingdom.

Population  A questionnaire was sent to 3074 ORACLE participants in a purposively selected sample of 55 collaborating maternity units, chosen to reflect a range of regions and of district general and teaching hospitals.

Methods  Content analysis was applied to verbatim text provided in response to an open question. Responses were also compared with a framework based on key points about the purpose of ORACLE. Closed questions were analysed using descriptive statistics.

Main outcome measures  Participants' interpretations of the purpose of the study.

Results  A response rate of 61% was achieved, and 1462 participants provided written answers to a specific question on why the study was being carried out. Content analysis suggested that the information leaflet was highly valued as a source of information about the trial. There was evidence that women's interpretations of the purpose of the trial were not identical to those that the investigators intended. Of the five key points about the trial described in the information leaflet, 400 (27%) participants reported one key point, 550 (38%) two key points, 229 (16%) three key points and 23 (1.5%) four key points. None reported five key points and it was not possible to classify 46 responses (3%). Vague, confused understanding or poor recall were evident in 204 (14%) of responses.

Conclusion  Although the ORACLE trial was run as a model of good practice at the time, this study suggests that it may not be possible to demonstrate full understanding of trial purpose and design by all participants. Emphasis should be on the provision of full information that involves consumers in its design and evaluation.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Little is known about the experiences of women recruited to clinical trials during pregnancy, particularly where those women have experienced a pregnancy-related emergency. The particular vulnerabilities of women in situations where they are dependent on clinical staff1 and may comply with doctors' and midwives' requests at such times, regardless of personal preference,2 have been recognised. Research with study participants has largely focused on general attitudes to trials and reasons for participation and non-participation.3–5 It is only recently that the importance of understanding the meanings given to trial explanations by participants has begun to be recognised.6 In this paper, we report an investigation into interpretations of study information by women participating in ORACLE, a randomised controlled trial of antibiotics in pregnancy.

The ORACLE trial was a double-blind placebo-controlled trial funded by the UK Medical Research Council.7,8 It was designed to test the hypothesis that treatment of women less than 37 weeks pregnant in either preterm labour or with prelabour rupture of membranes with broad-spectrum antibiotics (Augmentin and/or Erythromycin) prolongs pregnancy and improves neonatal mortality and morbidity. ORACLE used a 2 × 2 factorial design, with four treatment possibilities: Augmentin 375 mg; Erythromycin 250 mg; both of these antibiotics; or both placebos. The trial medicines were intended to be taken four times daily for 10 days or until delivery. During the period of the trial (July 1994–May 2000), 11,154 women were randomised to ORACLE from 161 maternity units worldwide, including 135 in the UK. Women were approached for recruitment to ORACLE on delivery suite when they were admitted in spontaneous preterm labour or with preterm rupture of the membranes.

The parents' (trial) information leaflet was developed prior to recruitment starting in July 1994, and was revised in 1995 and 1997 to reflect the changes in the evidence in this area. The National Childbirth Trust and the Association for Improvements in Maternity Services (both consumer groups in maternity care) were closely involved in the development of the leaflet. The final version was piloted in autumn of 1997 in 10 maternity units among staff and women. Between 1997 and 2000, ORACLE had 80 part-time midwives in the maternity units with over 3000 deliveries annually. These midwives underwent intensive training regarding the trial. The importance of informed consent was emphasised throughout.

A questionnaire was developed to investigate women's views about the trial information, their interpretation of the study and their adherence to study medicines. This paper focuses on women's interpretation of information.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

The questionnaire was extensively piloted, as was the letter used to accompany the questionnaire. Reminder letters were used to improve the response rate. The responses to one question (in relation to feeling under pressure to join the trial) was cause for concern and was followed up by letter with the relevant participants.

Participants

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

A sample of 55 maternity units recruiting women to ORACLE was purposively chosen to reflect a range of Regions and of District General and Teaching Hospitals. Ethical approval was given in each of the units for the questionnaire to be sent to women who had been randomised in ORACLE. The questionnaire was sent to women two weeks after they had joined the trial.

Analysis

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

The main focus of this paper is in responses to this question in the questionnaire:

Did anyone tell you why the study was being carried out? (Yes/No). If yes, could you put down in your own words what you were told?

Content analysis, a systematic method for assigning text to content categories,9 was used to analyse the verbatim data given in response to this question. The categories were generated through close inspection and comparison of the data. Explicit specifications were devised to guide data assignment which was assisted by QSR NUD*IST (ver 4) software. Two coders were used, and text searches were conducted to enhance the reliability and reproducibility of the coding. The themes of the analysis are illustrated below using representative quotations.

In addition, we assessed participants' responses when compared with a framework of key points about what the study involved as described in the trial information leaflet. Five key points about the trial relevant to participant interpretation were defined and assessed, as shown in Table 1.

Table 1.  Key points about why the ORACLE trial was undertaken.
1) There is a hypothesis that underlying cause of the labour or broken waters may be subclinical (silent) infection.
2) Antibiotics (not just ‘tablets’ or ‘medicine’) were being tested in the trial.
3) There is a hypothesis that treatment with antibiotics might prolong the pregnancy/delay delivery.
4) There is a need to look at babies' health and development.
5) Randomisation and blinding were necessary to evaluate these questions.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

A response rate of 61% (1875/3074 questionnaires completed) was achieved. Responders and non-responders appeared to have similar characteristics in terms of the proportion of women who had preterm labour, prelabour rupture of the membranes or both, and gestational age at recruitment, but no further comparisons between responders and non-responders were possible.

Information about the study

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Of the 1875 women who completed the questionnaire, 1633 (87%) stated that they were given the parents' (trial) information leaflet and that they had read it at the time (1467; 78%). Five reported that they had neither seen nor read the information leaflet nor talked to anyone. Forty-four (2%) women reported that they had seen the leaflet and read it at the time, but that no one had talked to them. The leaflet was identified by 797 (43%) of women as most helping them to understand the trial, compared with 565 (30%) who identified talking to the midwife, and 348 (19%) who identified talking to the doctor. Data were missing for 165 (8%) of women in response to this question.

In responding to a specific question, 266 (14%) women reported that they felt under pressure to join the study. When further explored by letter to these women it was found the majority of women perceived the pressure to arise from their circumstances (including pressures of time and worry) and not staff. Although we did not follow them up, 62 (4%) women reported that they felt under pressure not to join the study.

Interpretation of the purpose of the study

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

A total of 1462 (78% of those who completed the questionnaire) wrote in their own words a response to the question:

Did anyone tell you why the study was being carried out? (Yes/No). If yes, could you put down in your own words what you were told?

We identified three themes in responses to this question: the purpose, the design and participation in the study.

Purpose of the study

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

The most common interpretation of the purpose of the study was that it was connected with stopping early labour. Of these, 628 (45%) mention the role of antibiotics, 143 (10%) mention tablets and 50 (3.5%) mention drugs.

To see if giving antibiotics to women with spontaneous ruptured membranes or early labour would prolong the length of time baby stayed in womb and therefore mature.

Twenty-two women (incorrectly) reported that ORACLE was the drug in the trial rather than the name of the trial.

I was told that if I had any infection the ORACLE could clear it up and help prolong my pregnancy.

Design of the study

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Most answers (1001, or 68% of those who answered the question) indicated that participants understood that the study involved testing a hypothesis. The majority of these referred to the role of antibiotics or medicine in delaying the onset of labour, sometimes with elaborations about infection.

To test whether antibiotics could delay premature labour, as there is the possibility that prem labour could be caused by infection either in the amniotic fluid or other sources.

There was, however, evidence of confusion about which hypothesis was being tested, with responses indicating a range of possible hypotheses that might be under consideration. Much of this confusion related to the possible effects of antibiotics on the health of the baby, and an apparent confusion between the effects on antibiotics in delaying premature labour and preventing infection in the baby.

To establish whether antibiotics could help a baby survive if it was born prematurely.

To see if taking antibiotics would help prevent premature babies from contracting certain types of illness/infections on delivery.

There were indications in the statements of 82 respondents (6% of those who answered the question) that they saw the purpose of the study as exploratory, mostly concerned with identifying the causes of preterm birth or preterm rupture of the membranes.

I was told the Study was being carried out because they were trying to find a reason for pre-Labour.

If I joined the study the results would eventually help find the cause in babies being born prematurely.

There was evidence in the statements of 209 respondents (14% of those who answered the question) that they understood that some form of randomisation, involving the use of ‘dummy’ tablets, was involved in the trial. Of these, 46 (3%) made explicit reference to the use of blinding in the study.

I was told that the study was a new one. It's aim was to find out if antibiotics were any use to prem. labour and unborn children and no-one knew whether you were taking the real ones or not.

Participating in the study

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Statements given by 31 respondents (2% of those who answered the question) suggested that they believed that there was no risk of harm to them or their baby as a result of participating in the trial, or they presumed that there was a benefit in participation.

She said it might help my unborn baby in the future but couldn't guarantee it but she could guarantee that it would give no harm to the baby.

The statements of 27 (2%) women emphasised the voluntary nature of their participation in the study. This was signalled by the use of phrases such as “I was asked if I would like to”.

I was told that taking part in the study was up to me. That it was good if I took part but again, the choice was mine.

An element of altruism, or wanting to help women in the future in a similar situation, was evident in the statements of 66 respondents (4.5% of those who answered the question) as a motivation for participating in the trial.

The midwife explained that you thought premature labour may be caused by an infection in the womb and by taking these tablets for a period of ten days—it may be proved your theory is true and therefore we have something in the future for women who have premature labour.

Interpretations of key points about the purpose of the study

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Participants' responses were assessed against a framework of five key points about why the trial was being carried out (Table 1). These responses are summarised in Table 2.

Table 2.  Reporting of key points about purpose of the study.
Assessment of verbatim responsesn (%)
Reported one key point400 (27)
Reported two key points550 (38)
Reported three key points229 (16)
Reported four key point23 (1.5)
Reported five key points0 (0)
Unclassifiable46 (3)
Assessed as vague or confused129 (9)
Reported no or poor recall75 (6)

Of the 400 (27%) who reported one key point, the emphasis was on delay in delivery of the baby.

To try to stop the baby coming early.

Of the 550 (38%) participants who reported two key points about the trial, the emphasis, once again, was on delay in delivery. References to antibiotics were commonly included.

That it was to determine whether or not the use of antibiotics could prevent premature labour.

Two hundred and twenty-nine (16%) of the participants reported three points and 23 (1.5%) of women reported four points (often including medical terminology).

Basically that is believed that women go into premature labour due to internal silent infection, therefore, by giving a course of antibiotics this may prevent this and reduce the risks involved in early labour.

No participant reported five points. We were unable to classify 46 (3%) responses in terms of key points because of issues such as brevity, difficulty in interpreting what was intended or no reference to key points.

All that we asked was answered.

We assessed 129 (9%) of responses as suggesting that participants' interpretation was vague or that they were confused about the purpose of the trial.

I was told it was because of research.

Seventy-five (6%) participants commented on their recall of the purpose of the trial, mostly to say that they could not remember. Twelve commented that the circumstances at the time impaired their memory of the explanation of the trial.

Very briefly and to be quite honest, I can't really remember, I was in quite a state as I was in premature labour at 30 weeks.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Obtaining consent to research from pregnant women in emergency situations raises important ethical issues and questions of best practice. The ORACLE trial was run as a model of good practice at the time. The trial information leaflet was welcomed by women in the study. The majority were able to report one key point about the purpose of the study, with a minority (14%) of answers demonstrating vagueness, confusion, or poor recall, and a further 3% unclassifiable. Our study suggests that even with high quality leaflet design, training of recruiting staff and provision of information, all participants are unlikely to be able to show that they have interpreted the purpose of the trial and the terms and concepts involved exactly as intended.

This study has its limitations. The questionnaire methodology chosen does not allow the researcher to ask for clarification nor the woman the opportunity to respond to questions not included,10 although we did attempt to address this in part by following up some participants by letter. It is also difficult to judge the extent to which written responses represent difficulties in written expression. The questionnaire was sent to women two weeks after they had joined ORACLE, thus relying on their ability to remember what they had understood. Other researchers report poor recall of labour events among women.11 Our study did not elicit the views of the doctors and midwives recruiting these women. In addition, although our response rate was acceptable for a study of this type, a considerable proportion of ORACLE participants did not respond.

Participants' responses suggested that 56% understood the purpose of the trial to be concerned with stopping preterm labour. This contrasts with the researchers' priorities, which were focused on whether neonatal mortality and morbidity were reduced. This finding suggests a possible divergence, although not necessarily a serious one, between the interpretations of participants and researchers in the research process. While it is of concern that 14% of participants report vagueness or poor recall of the purpose of the trial, this may not be surprising. Consent was obtained in unfamiliar circumstances where women are likely to experience extreme stress and anxiety. It is reassuring that only 2% of participants reported spontaneously that they were believed that there was no possible harm associated with the study, although a direct question might have elicited a different response. It is similarly of interest that only a small proportion (4.5%) identify altruism as motivating their participation, although again a direct question might have provoked different answers.

The information leaflet was rated by more participants as being helpful than either doctors or midwives. It is possible that the leaflet could have been further improved by more directly involving women who had been through the experience of preterm labour.6 However, it is also important that some limits to what can be achieved by improving provision of information are recognised. ORACLE did comply with best practice at the time, and there was a strong emphasis on the quality of the consent process. Even with these efforts, our findings confirm recent research suggesting that the meanings that participants give to research may not be precisely the same as those that researchers intend them to have.12 These findings have important implications for what is realistic to achieve in terms of informed consent. It is important that recent government initiatives13,14 on patient information should not be dominated by a mechanistic approach,15 where the process is judged by the extent to which participants demonstrate ‘understanding’. It may not be possible for all participants in research to show full and complete understanding of the scientific issues of each study, or even an understanding of the general principles underlying trial design. A recognition of this may lead to a reformulation our ideas on informed consent, so that the emphasis is on process—the provision of full information that involves consumers in its design and evaluation, rather than outcome—the expectation that all participants can demonstrate that they are fully informed.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

The authors would like to thank all the women who completed the questionnaire, the ORACLE Trial staff: Rebecca Smyth, Amanda Rhodes, Ann Blackburn, Kate Taylor, Fran Vickers, Jayne Matlock and Nadine Gilbert. Thanks also to David Taylor, Peter Brocklehurst and Ellen Carter.

Contributions

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

SK conceived the study, conducted the literature review, devised the questionnaire, and supervised its administration. SK and MDW devised the coding scheme. SK and MDW both undertook coding. Both authors drafted the paper.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. Participants
  6. Analysis
  7. RESULTS
  8. Information about the study
  9. Interpretation of the purpose of the study
  10. Purpose of the study
  11. Design of the study
  12. Participating in the study
  13. Interpretations of key points about the purpose of the study
  14. DISCUSSION
  15. Acknowledgements
  16. Contributions
  17. Funding
  18. References

Accepted 10 May 2004