EMPIRICAL EVALUATION OF THE NEED FOR ‘ON-GOING CONSENT’ IN CLINICAL RESEARCH

Abstract

Objective

Some commentators argue that informed consent for clinical research should be an on-going process, which begins, rather than ends, with participants’ initial consent. Lacking, however, are empirical data on whether there is a need for ‘on-going consent’.

Design

Two self-administered surveys—a baseline survey at initial consent and a follow-up survey 2-3 years later—to assess whether participants remain informed over time.

Methods

Respondents were adults with HIV disease from Argentina, Brazil and Thailand enrolled in a long-term clinical trial.

Results

Respondents overall were well informed at baseline. At follow-up, many reported being not informed about aspects of the study central to their on-going participation. With respect to the possibility of withdrawal, 38.5% of respondents at follow-up reported being “not at all” informed. At follow-up, 71.1% wanted more information. Yet, 62.8% had not asked any questions during the entire study. Reasons for not asking questions included not having an opportunity (16.4%) and not knowing whom to ask (15.5%).

Conclusions

The standard consent process resulted in participants being well informed at enrollment. Yet, this process was not sufficient to keep them informed about aspects of the study central to their on-going participation. In addition, participants who wanted more information often did not ask for it. These findings provide empirical support for recommendations that clinical trials should consider including a process of ‘on-going consent’.

Introduction

Informed consent has been recognized as an ethical requirement for clinical research since at least the Nuremberg Code of 1947 [1,2]. Standard practice implements this requirement in four steps: explaining the study to potential participants, answering any questions, obtaining their informed consent, and providing them with a copy of the consent form [3,4]. Yet, some commentators argue that this standard consent process is not sufficient to fully protect research participants [5,6]. They argue that informed consent for clinical research should include a process to keep participants informed over time about aspects of the study relevant to their on-going participation [7-13].

Recommendations for what has been called ‘on-going consent’ make theoretical sense [14]. Participants may forget relevant information over the course of their research participation, such as the right to withdraw. Changes also may occur after participants have enrolled: new options may become available, participants’ health may decline, their diseases may progress [15].

These theoretical possibilities highlight the need for empirical data to evaluate whether, in practice, the standard consent process is sufficient for clinical research [16]. Specifically, is the standard informed consent process sufficient to keep research participants informed over time about aspects of the study relevant to their on-going participation? To begin to address this question, the present study evaluated the understanding and desire for information of individuals who were participating in a longer-term clinical trial.

Methods

Participants

The present data were collected from two self-administered surveys of individuals participating in the Evaluation of Subcutaneous Proleukin in a Randomized International Trial (ESPRIT). ESPRIT was an open-label trial which compared antiretroviral therapy and interleukin-2 (IL-2) to antiretroviral therapy alone in individuals infected with HIV [17,18]. The study enrolled 4150 HIV-positive men and women, 18 years or older, with a minimum baseline CD4 count of 300/μL in 25 countries, mostly in Europe. Enrolled participants returned to the clinic every 4 months for medical evaluation and collection of blood samples. Participants randomized to IL-2 received three cycles of subcutaneous IL-2 during 5 consecutive days every 8 weeks, and then additional cycles based on their CD4+ cell count. Median follow-up in ESPRIT was 6.8 years.

Concern has been expressed that research participants in developing countries are less likely to provide valid informed consent. With this in mind, the present survey focused on the 3 developing countries that participated in ESPRIT: Argentina, Brazil, and Thailand.

ESPRIT Consent Procedures

ESPRIT thoroughly implemented the standard consent process used in clinical research. In Argentina and Brazil, potential participants had an initial meeting with an ESPRIT researcher who explained the entire study, went over the consent form, and answered questions. In Thailand, the initial meeting involved a 2-hour group discussion led by research nurses in which the consent form was read page by page, and discussed with the group. In all 3 countries individuals were then given time to consider the study and discuss it with others, including family and clinicians. At a second meeting, potential participants met with a researcher who again explained the study and answered any questions prior to inviting interested individuals to sign the consent form.

The consent form was 2,600 words long and described all the essential elements of informed consent. With respect to the possibility of withdraw from the study, the consent form stated: “Your participation is entirely voluntary. You may decide not to take part or to withdraw from the study at any time without losing the benefits of your routine medical care.” With respect to the post-trial provision of IL-2, the consent form stated: “If the research finds that IL-2 is safe and effective for HIV patients, and you and your doctor decide that you want to take it, the company supplying IL-2 during this study will provide it to you until it is approved for use for HIV infection.” All participants were provided with a copy of the consent form. Enrolled participants also received several letters during the study which informed them of new information regarding the side effects of IL-2 and extension of the study.

Surveys

ESPRIT participants in Argentina, Brazil, and Thailand were invited to participate in two self-administered surveys, based on the availability of clinic staff. The baseline survey was completed just after consent, but prior to randomization. Individuals were invited to participate in the follow-up survey after they had been enrolled in ESPRIT for 2-3 years, whether or not they had completed the baseline survey. Completed surveys were mailed to the statistical and data management center for ESPRIT at the University of Minnesota; clinical research staff did not have access to the completed surveys.

The surveys were developed in 9 steps: 1) literature review; 2) draft survey development; 3) review and revision by survey professionals; 4) cognitive, behavioral, and reliability pretesting with ESPRIT participants in the United States; 5) revision; 6) translation into Portuguese, Spanish, and Thai; 7) back-translation into English and revision; 8) translation into Portuguese, Spanish, and Thai; and 9) review of the translated surveys by native investigators.

The two surveys used primarily close-ended questions in 4 domains: 1) understanding of ESPRIT; 2) motivations for participation; 3) impact of participation; and 4) compliance with study requirements. The present paper reports findings from the first domain. Responses in this domain were examined to assess participants’ understanding of IL-2 and ESPRIT, whether they wanted more information, whether they had asked any questions during the study, and whether they felt any pressure to remain in the study (see Appendix 1 for the verbatim questions).

Statistical Analysis

The baseline characteristics of respondents were compared using chi-square analysis for categorical data and the Kruskal-Wallis test for continuous data. For questions with ordinal responses on the follow-up questionnaire, a proportional odds model was used to assess the effect of treatment group after adjustment for gender, age, race, country, and participation in the baseline survey. Logistic regression was used for binary responses. For respondents who completed both the baseline and follow-up surveys, responses were compared by time period, with and without a term for treatment group by time interaction, using generalized models appropriate for multinomial paired data. The interaction in these models assesses whether the difference in responses between baseline and follow-up varies by treatment group.

Approvals

This study was approved by the Dr Virgilioss G. Foglia Ethics Committee, Buenos Aires, Argentina; the Emilio Ribas Institute of Infectious Disease Ethics Committee, Sao Paulo, Brazil; the Ethics Committee of the Faculty of Medicine, Chulalongkorn Hospital, Bangkok, Thailand; and the institutional review board at the National Institute of Allergy and Infectious Diseases, Bethesda, Maryland. All respondents provided informed consent before completing the surveys.

Results

Respondents

The participating sites in the 3 countries enrolled 854 participants in ESPRIT. The baseline survey was completed by 318 participants (37.2%) just after consent, but prior to randomization (Figure 1). While formal response rates were not collected for the baseline survey, investigators report that only a few invited participants declined to complete the baseline survey. Of the 854 ESPRIT participants at the sites in the 3 countries, 595 (69.7%) were invited to complete the follow-up survey, and 582 (response rate= 97.8%) did so. Of these 582 participants, 292 were receiving IL-2 and 290 were in the control arm; 253 of those who completed the follow-up survey had also completed the baseline survey.

Figure.

Subject Selection

Respondents to the follow-up survey had been enrolled in ESPRIT for a mean of 2.5 years. Nearly 70% were male, mean age was 37.6, approximately 8 in 10 were employed, and approximately 40% had attended college or university (Table 1). The characteristics of ESPRIT participants who did not complete the follow-up survey, ESPRIT participants who completed both the follow-up and baseline surveys, and ESPRIT participants who completed the follow-up survey only were similar (Table 4).

Table 1.

Characteristics of Respondents to the Baseline and Follow-up Surveys

	BASELINE SURVEY			FOLLOW-UP SURVEY
	Argentina (n = 168)	Brazil (n = 32)	Thailand (n =118)	Argentina (n = 351)	Brazil (n =43)	Thailand (n =188)
Male (%)	76.8	75.0	52.5	77.5	74.4	50.5
Female (%)	23.2	25.0	47.5	22.5	25.6	49.5
National health coverage (%)	59.5	46.9	8.5	50.4	83.7	10.6
Employer Insurance (%)	22.0	12.5	12.7	25.1	0.0	16.0
Self pay (%)	3.6	28.1	70.3	4.3	2.3	58.5
Other (%)	7.7	3.1	7.6	12.8	0.0	2.7
Insurance not reported (%)	7.1	9.4	0.8	7.4	14.0	12.2
No schooling (%)	1.2	0.0	0.0	0.9	2.3	0.0
< 12 years schooling (%)	26.8	3.1	28.8	20.2	25.6	25.0
High school graduate (%)	45.8	43.8	22.9	34.2	30.2	32.4
College or university (%)	20.2	37.5	32.2	21.1	34.9	36.2
Graduate school (%)	3.6	6.3	15.3	20.5	7.0	4.3
Schooling not reported (%)	2.4	9.4	0.8	3.1	0	2.1
Employed (%)	59.5	59.4	76.3	78.6	65.1	80.9
Employment not reported (%)	4.8	9.4	0.8	1.7	2.3	0.5
Age (mean years)	38.1	42.3	36.8	38.1	40.3	36.0
CD4+ count (cells/mm3) at baseline (median)	422	417	391	446	422	402

Table 4.

Comparison of baseline characteristics of ESPRIT participants who did not complete the follow-up survey, those who completed the follow-up survey but not the baseline survey, and those who completed both surveys

Baseline Characteristics	All ESPRIT participants	ESPRIT participants who did not complete follow-up survey	ESPRIT participants who completed follow-up but not baseline survey	ESPRIT participants who completed follow-up & baseline survey
No. patients	854	272	329	253
Age (median years)	37	37	36	36
Female (%)	31.3	30.9	29.2	34.4
Baseline CD4 (median cells/mm3)	475	484.5	482.7	455.9
IL-2 treatment group (%)	49.8	48.9	47.1	54.2

Are Participants Informed at Follow-up?

Respondents to the follow-up survey reported being significantly less informed than respondents to the baseline survey. This was evident in all those who completed the follow-up survey (Table 2) and also when the analysis was restricted to those who completed the baseline and follow-up survey (Table 3).

Table 2.

How Informed Participants Reported Being at Baseline) and at Follow-Up (mean=2.5 Years Later) (%)^#

			Very Well Informed	Somewhat Informed##	Not at All Informed	p- value*
Side Effects	Initial (n=300)		68.3	31.0	0.7
	Follow-Up	IL-2 (n=267)	76.0	19.9	4.1	< .001
		Control (n=245)	35.5	44.9	19.6
Benefits	Initial (n=294)		77.6	22.1	0.3
	Follow-Up	IL-2 (n=265)	69.1	26.4	4.5	< .001
		Control (n=241)	38.2	41.1	20.7
Randomization**	Initial (n=288)		63.9	33.3	2.8
	Follow-Up	IL-2 (n=258)	39.1	36.4	24.4	< .001
		Control (n=251)	26.3	34.3	39.4
Withdrawal+	Initial (n=291)		68.4	28.9	2.7
	Follow-Up	IL-2 (n=151)	35.8	29.1	35.1	.15
		Control (n=148)	26.4	31.8	41.9
Post-Trial Provision++	Initial (n=293)		48.1	46.1	5.8
	Follow-Up	IL-2 (n=256)	31.3	34.4	34.4	.043
		Control (n=241)	21.6	36.5	41.9
Frequency of IL-2 Injections	Initial (n=297)		79.5	20.2	0.3
	Follow-Up	IL-2 (n=282)	62.8	28.7	8.5	< .001
		Control (n=261)	34.5	36.4	29.1

^#Because respondents could skip questions the denominators vary across questions.

^##These data include those who answered “moderately” or “slightly” informed in the baseline survey and those who answered “somewhat” informed in the follow-up survey.

^*Comparison of IL-2 arm versus controls at follow-up based on multivariate logistic regression adjusting for gender, age, race, country, and participation in baseline survey. Other predictors were: Injection frequency: country, baseline cohort; Randomization: baseline cohort; Side effects: gender, baseline cohort; Benefits: gender, baseline cohort; IL-2 provision at study end: baseline cohort; Possibility of withdrawal: baseline cohort.

^**This question in the baseline survey was worded as “The chance of getting IL-2” and in the follow-up survey as “The percent of people enrolled in ESPRIT who get IL-2”.

⁺This question appeared only on the surveys in Argentina.

⁺⁺This question in the baseline survey was worded as “How you might get IL-2 after the ESPRIT study was over” and in the follow-up survey as “Whether the ESPRIT sponsors will provide IL-2 to ESPRIT participants after the study ends.”

Table 3.

How Informed are Participants who Completed both Surveys (%)

Question		N	Very well informed	Somewhat informed	Not informed	P- value*	Inter- action
Injection frequency							< .001
IL-2	Baseline	127	81.9	17.3	0.8	.006
	Follow-up	135	67.4	28.9	3.7
Control	Baseline	109	82.6	17.4	0.0	< .001
	Follow-up	107	38.3	41.1	20.6
Randomization							.72
IL-2	Baseline	122	71.3	25.4	3.3	< .001
	Follow-up	126	43.7	37.3	19.0
Control	Baseline	106	58.5	40.6	0.9	< .001
	Follow-up	105	32.4	37.1	30.5
Common side effects							< .001
IL-2	Baseline	127	73.2	26.0	0.8	.26
	Follow-up	130	78.5	20.0	1.5
Control	Baseline	111	63.1	36.9	0.0	< .001
	Follow-up	102	41.2	48.0	10.8
Potential benefits							.12
IL-2	Baseline	126	81.0	18.3	0.8	.46
	Follow-up	128	70.3	25.0	4.7
Control	Baseline	107	78.5	21.5	0.0	.003
	Follow-up	101	47.5	41.6	10.9
IL-2 provision at study end							.091
IL-2	Baseline	125	56.0	40.0	4.0	< .001
	Follow-up	125	32.8	36.8	30.4
Control	Baseline	106	41.5	52.8	5.7	.002
	Follow-up	101	31.7	42.6	25.7
Possibility of withdrawal							.74
IL-2	Baseline	125	67.2	31.2	1.6	< .001
	Follow-up	62	43.5	29.0	27.4
Control	Baseline	108	66.7	30.6	2.8	< .001
	Follow-up	50	36.0	38.0	26.0

^*P-value is for differences by time period.

^**Time by treatment group interaction p-value shown in italics.

With respect to the possibility of withdrawal, 2.7% at baseline but 38.5% at follow-up (IL-2 35.1%; control 41.9%) reported being “not at all” informed. With respect to the possibility of receiving IL-2 after the study, 5.8% of those at baseline but 38.0% of those at follow-up (IL-2 34.4%; control 41.9%) reported being “not at all” informed (Table 2). Not surprisingly, respondents in the IL-2 arm were significantly more likely to report being well informed at follow-up about IL-2 (Table 3).

Wanting Information

Of respondents to the follow-up survey, 71.1% (67.7% IL-2; 74.5% control) responded “yes” to the question: “At present, would you like any additional information about the ESPRIT study”. Of these, 27.1% (31.7% IL-2; 23.0% control) wanted more information about the possibility of withdrawal, 61.4% (66.1% IL-2; 57.1% control) wanted more information about the possibility of receiving IL-2 after ESPRIT ended, and 72.6% (73.9% IL-2; 71.4% control) wanted to know when they would learn about study results. Respondents in the IL-2 arm were significantly more likely to want more information about the possibility of withdrawing from ESPRIT (p= .029).

Obtaining Information

Sixty-five percent of respondents to the follow-up survey (61.6% IL-2; 67.8% control) reported still having a copy of the consent form; another 14% (15.1% IL-2; 13.0% control) could not recall if they still had a copy. Fifty-eight percent (59.7% IL-2; 55.6% control) had looked back at the form at least once; 12.4% (13.6% IL-2; 11.3% control) had looked back at the form at least 3 times. By the time of follow-up, 2-3 years after enrollment, 62.8% of respondents (59.0% IL-2; 66.8% control) had not asked the research team any questions. Of those who did not ask any questions, 16.4% (9.5% IL-2; 22.6% control) reported not having an opportunity, 15.5% reported not knowing whom to ask (13.9% IL-2; 16.9% control), 8.1% reported not wanting to be removed from the study (7.6% IL-2; 8.5% control), and 6.9% reported not wanting to look foolish (7.6% IL-2; 6.2% control).

Forty-seven percent of respondents to the follow-up survey (51.1% IL-2; 43.4% control) reported receiving information about HIV treatment from sources other than the ESPRIT team. The most frequently cited sources were personal doctors (67.1%), newspapers and magazines (38.1%), and television (33.3%). Seventy-two percent (72.1% IL-2; 72.0% control) found the information from outside sources added to what they had learned from the ESRIT team; 0.9% reported that the information conflicted with what they had learned. Sixty-one percent (60.6% IL-2; 61.1% control) reported that this information made them more enthusiastic to participate; 3.3% reported that the information made them more hesitant to participate.

Pressure

Of respondents to the follow-up survey, 19.4% (22.4% IL-2; 16.4% control) responded “a great deal”, “a moderate amount”, or “a little” to the question “How much pressure, if any, do you feel from other people (including the ESPRIT research team) to continue in the ESPRIT study”. The most common sources of pressure were family/close friends (42.5% -- 38.7% IL-2; 47.7% control), the ESPRIT team (36.8% -- 41.9% IL-2; 29.5% control) and the respondent’s personal doctor (23.6% -- 27.4% IL-2; 18.2% control). Treatment group, gender, age, race, country, and participation in the baseline survey were not significantly associated with feeling pressure to continue to participate in ESPRIT.

Discussion

The present data provide the first empirical evaluation of which we are aware on whether the standard consent process used in clinical research is sufficient to keep participants informed over time. The primary finding is that the standard consent process did not keep participants of ESPRIT in Argentina, Brazil, and Thailand informed about aspects of the study central to their on-going participation. Most importantly, at follow-up, almost 4 in 10 respondents reported being “not at all” informed about their right to withdrawal, and almost 4 in 10 reported being “not at all” informed about the possibility of receiving the experimental treatment after study completion.

These findings are striking given the context in which they were obtained. ESPRIT’s implementation of the standard consent process was comprehensive and in-depth. The effectiveness of this process was assessed in a separate empirical study that evaluated the actual understanding of ESPRIT participants in Thailand [19]. This testing found that, immediately following the initial consent process, participants in Thailand were well informed.

The fact that ESPRIT thoroughly implemented the standard consent process, and ESPRIT participants were well informed at enrollment, suggests that the present findings provide an assessment of the standard consent process under a best case scenario. In particular, it seems extremely unlikely that the present participants’ desire for more information traces to their having undergone a sub-optimal implementation of the standard consent process at baseline. Instead, these findings suggest that the standard consent process may not be sufficient to protect fully the participants of clinical trials. This conclusion provides empirical support for the claim that longer-term clinical trials should consider including a process to ensure that participants remain informed over time about aspects of the study relevant to their on-going participation, such as the right to withdraw.

It is important to distinguish what has been called ‘on-going consent’ from initial consent and re-consent. Initial consent is designed to provide a comprehensive explanation so potential participants can decide whether to enroll in a given study. Re-consent is designed to provide enrolled participants with all the information they need to decide whether to continue to participate given significant changes, such as identification of serious new risks [20]

On-going consent, in contrast, is intended to keep participants aware of information relevant to their continued participation. On-going consent occurs subsequent to participants’ initial enrollment, and in settings where no substantial changes have occurred to the study that would warrant obtaining their reconsent. This suggests that on-going consent need not use a consent form nor obtain participants’ signature. In addition, a good deal of the information needed for initial consent is not germane to on-going consent. Hence, on-going consent can be much briefer than initial consent, and should cover a much narrower range of information.

For example, at the time of the follow-up survey, respondents in the control arm did not feel informed about IL-2. While this information was important to initial consent, it seems unimportant to the on-going participation of individuals who have been randomized to the control arm. In contrast, awareness of the risks and potential benefits of IL-2 could be crucial to these individuals’ options following the study. This possibility suggests that on-going consent might include an exit interview at the completion of individuals’ participation to provide them with information relevant to their next steps.

On-going consent could be implemented by incorporating into clinical trials periodic and informal discussions. In particular, investigators might briefly discuss: 1) up-coming aspects of the study; 2) any changes to participants’ circumstances or the study; 3) the importance of the study; 4) the importance of continued participation; and 5) participants’ right to withdraw.

A process of on-going consent would provide research participants an explicit opportunity to ask questions. This opportunity seems important given the present findings that almost 2 in 3 respondents wanted more information, yet over half of these participants had not asked any questions, often because they did not know whom to ask or felt that they did not have an opportunity to ask questions.

The content of the on-going consent meetings should be tailored to the circumstances. What information needs to be highlighted will vary depending on the specific study and participants’ point in the study. The process of on-going consent also should be sensitive to the cultural context. For example, the best approach to providing participants sufficient opportunity to ask questions likely will vary across cultures and countries.

Almost half of the respondents to the follow-up survey had obtained information from outside sources. A few reported that this information conflicted with what the research team had told them, and made them less enthusiastic about participation. While this effect may be especially prominent in research on HIV, for which there are many sources of information,inclusion of a process of on-going consent would provide all investigators with an opportunity to address the possibility of participants receiving inaccurate information.

Respondents who had completed the baseline survey reported being significantly better informed at follow-up, 2-3 years later. While it is unclear what explains this finding, future research might evaluate whether it is possible to supplement the initial consent process in ways that help participants retain relevant information over time.

Some respondents reported feeling pressure from the team to remain enrolled. This finding, while important, should be interpreted with caution. Reports of “pressure” may reflect a range of experiences, not all of them problematic. Reminding participants of the importance of their participation and even encouraging them to continue is appropriate provided it is done respectfully and they are assured of their right to withdraw. The fact that some respondents felt pressure to continue to participate from their family/close friends or personal doctor suggests investigators also should be cognizant of other potential sources of pressure.

The present findings are subject to four important limitations. First, the findings are based on self reports. While individuals’ perceptions may not always track how well informed they in fact are, the present respondents’ reports of being well informed at baseline correlate with data showing that ESPRIT participants in Thailand were well informed at initial consent [19]. Second, the present data were collected in 3 countries from sites of a single study for HIV disease. Results may differ for other diseases, and in other places. Third, the right to withdraw [21] and post-trial provision of experimental treatments can be difficult concepts to understand. However, the consent form for ESPRIT seemed to explain both items well. Fourth, the present data do not evaluate the impact on clinical research trials of incorporating a process of on-going consent. Before on-going consent is endorsed generally, pilot testing should be conducted to evaluate its feasibility, cost, and impact on clinical trials.

Appendix: Verbatim Questions

A. Are Participants Informed at Follow-up?

Item	Initial Survey	Follow-up Survey
	“How well informed do you think you are about…?	Please indicate how well informed you currently feel about each of the following aspects of the ESPRIT study
a. Side Effects	The common side effects of IL-2	The common side effects of IL-2
b. Benefits	The potential benefits of getting IL- 2	The potential medical benefits of IL-2
c. Randomization	The chance of getting IL-2	“The percent of people enrolled in ESPRIT who get IL-2
d. Withdrawal	“The possibility of withdrawing from the study	The possibility of withdrawing from the ESPRIT study
e. Post-Trial Provision	How you might get IL-2 after the ESPRIT study is over”	“Whether the ESPRIT sponsors will provide IL-2 to ESPRIT participants after the study ends
f. Frequency of IL-2 Injections	The number of IL-2 injections and clinic visits	“How often ESPRIT participants give themselves IL-2 injections

Answer choices for the initial survey were: Very Informed; Moderately Informed; Slightly Informed; Not informed at all. Answer choices for the follow-up survey were: Very well informed; Somewhat informed; Not informed at all.

B. Wanting Information at Follow-up

1. Please indicate which aspects of the ESPRIT study you wish you had received MORE information about when you first enrolled in the ESPRIT study.

1. How often ESPRIT participants give themselves IL-2 injections.

2. The percent of people enrolled in ESPRIT who get IL-2.

3. The common side effects of IL-2.

4. The potential medical benefits of IL-2.

5. Whether the ESPRIT sponsors will provide IL-2 to ESPRIT participants after the study ends.

6. The possibility of withdrawing from the ESPRIT study.

Answer choices were: I wish I had received a lot more information; I wish I had received a little more information; I did not need more information.

C. Obtaining Information During the Study

1. Since you began participating, have you ever asked the ESPRIT research team for any additional information about the study?

Answer choices were: Yes; No (these respondents were instructed to skip the next question).

2. Why haven’t you asked the ESPRIT research team for any additional information about the study? (Mark ALL that apply)

Answer choices were: I didn’t want any additional information; I didn’t know who to ask; I didn’t want to look foolish; I didn’t want to be removed from the study; I didn’t have an opportunity to ask; Other: Please specify (open-ended response)

D. Retaining Consent Form

1. Do you still have a copy of the ESPRIT consent form?

Answer choices were: Yes; No; Don’t Know (these respondents were instructed to skip the next question)

2. Since the time you signed the ESPRIT consent form, how many times have you looked back at it?

Answer choices were: Never; 1-2 times; 3-5 times; 6-10 times; More than 10 times.

E. Asking Questions

1. Since you began participating, have you ever asked the ESPRIT research team for any additional information about the study?

Answer choices were: Yes; No (these respondents were instructed to skip the next question).

2. Why haven’t you asked the ESPRIT research team for any additional information about the study? (Mark ALL that apply)

Answer choices were: I didn’t want any additional information; I didn’t know who to ask; I didn’t want to look foolish; I didn’t want to be removed from the study; I didn’t have an opportunity to ask; Other: Please specify (open-ended response)

F. Obtaining Information from Outside Sources

1. Since you began participating in the ESPRIT study, have you received information about HIV treatments from sources other than the ESPRIT research team?

Answer choices were: Yes; No (these respondents were instructed to skip the following 3 questions).

2. What was the most useful source of information?

Answer choices were: Your personal doctor; A hospital or health clinic (other than your ESPRIT clinic); Other people already taking IL-2; HIV community or activist information networks; Newspapers or magazines; Television; World-Wide-Web/Internet; Other sources: Please specify (open-ended response)

3. Overall, compared to what you learned about HIV treatments from the ESPRIT research team, did the additional information you received from other sources…

Answer choices were: Add to what you learned about HIV treatments from the ESPRIT research team; Repeat what you learned about HIV treatments from the ESPRIT research team; Conflict with what you learned about HIV treatments from the ESRIT research team

4. What impact, if any, did the additional information have on your willingness to continue participating in the ESPRIT study?

Answer choices were: It made me much more enthusiastic about continuing to participate; It made me somewhat more enthusiastic about continuing to participate; It had no impact on my willingness to continue to participate; It made me somewhat more hesitant about continuing to participate; It made me much more hesitant about continuing to participate.

G. Pressure

1. How much pressure, if any, do you feel from other people (including the ESPRIT research team) to continue in the ESPRIT study?

Answer choices were: A great deal; A moderate amount; A little; None (these respondents were instructed to skip the next question).

2. From whom did you feel pressure? (Mark ALL that apply)

Answer choices were: Your personal doctor; The ESPRIT research team; Your family member(s); Your close friend(s); Other ESPRIT participants; Others: Please specify (open-ended response).