Annals of Oncology Advance Access published online on October 15, 2008
Annals of Oncology, doi:10.1093/annonc/mdn623
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The length of consent documents in oncological trials is doubled in twenty years
1 Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, St Olavs Hospital, Trondheim
2 National Resource Centre for Studies of Long-term Effects after Cancer, Rikshospitalet University Hospital, Oslo
3 Department of Oncology, St Olavs Hospital (Trondheim University Hospital), Trondheim, Norway
* Correspondence to: Mr. O. Berger, Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, St Olavs Hospital, N-7005 Trondheim, Norway. Tel: +4797519419 Fax: +4773867289; E-mail: olabe{at}stud.ntnu.no
 |
Abstract |
|---|
 Top Abstract introductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
Background: The aim of the study was to investigate whether the length of informed consent documents (ICDs) for oncological trials have increased from 1987 to 2007 and analyze the content of the ICDs.
Design: In total, 87 ICDs from oncological trials approved by the Regional Committee for Medical and Health Research Ethics (REC) in the central region of Norway from 1987 to 2007 were analyzed. A list of 17 basic (fundamental medical and ethical aspects) and 30 formal (juridical aspects, financing, insurance and storage of data) content components was constructed based upon international and REC guidelines for ICDs. The number of words and presence of components were registered for all ICDs.
Results: The mean length of the ICDs increased from 338 (range 276–464) words in 1987–1990 to 1087 words (range 399–2345) in 2005–2007. The number of components increased from nine to 25 during the same period. Basic components increased steadily from seven in 1987–1989 to 14 in 2005–2007 while the components concerning formalities increased substantially from two to 11.
Conclusions: The increased length of the ICDs is explained by an increased complexity of the documents and especially more information about formalities. This development increasingly demands competent readers and might prohibit truly informed consents.
cancer, clinical trial, content, informed consent, length
|
introduction |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
Prior to World War II, ethics in medical research was poorly defined and experiments were carried out in ways that now would be considered unethical. The medical ‘experiments’ carried out during World War II aroused the international community's awareness of unethical medical research. As a consequence, the Nuremberg Code with its principal guidelines for ethics in research on human beings, was published in 1947 [1, 2].
With a few exceptions, e.g. unconscious patients or patients with dementia, clinical research today presupposes that patients participate voluntarily. This is based on a fundamental respect for the patients' autonomy, dignity, integrity and self-determination. To ensure the patients' rights in clinical research, specific ethical rules and guidelines have been worked out on how to give sufficient information before enrolling patients in a study [3]. Oral information from health care professionals is considered by many clinicians to be an important part of this process [4, 5]. The written informed consent document (ICD) is, however, mandatory and essential because it standardizes the information, regulatory authorities may influence upon the content and one can secure that all patients receive all necessary and relevant information. Furthermore the patient has a copy of a document with information that could be used as a reference during the study participation.
The Declaration of Helsinki states fundamental requirements to what ICDs should contain [6]. The ICDs are to include information that participation is voluntary, about aims, methods, sources of funding, any possible conflicts of interest, anticipated benefits and any potential risks of participating in the study. Further it emphasizes that patients should not give their written informed consent before understanding this information. The Council for International Organizations of Medical Sciences (CIOMS), a part of World Health Organization, has since its establishment in 1949 published international guidelines with revised recommendations in 1982, 1993 and 2002 on how the requirements in the Declaration of Helsinki should be met [7–9].
Since the establishment of the the Regional Committees for Medical and Health Research Ethics (RECs) in Norway in 1986 there has been a steady increase in international and national recommendations on the content of the ICDs [8, 9]. The RECs are responsible for controlling that the shape, length and content of the ICDs are in accordance with The Declaration of Helsinki and the CIOMS's recommendations [10]. Research protocols must also be approved by other regulatory authorities separately. Depending on the type of project, protocols and ICDs must be accepted by The Norwegian Medicines Control Authority [11], The Ministry of Health and Social Affairs [3], The Norwegian Biobank Committee [12] and The Data Inspectorate [3]. All have claims about what information has to be included in the ICDs such as information about insurance, possibilities of compensation in case of injury, information about who is responsible for the trials, economical matters and storing and publishing of data.
In 1989, LoVerde et al. [13] studied 88 consecutive research consent documents from the Denver Veterans Administration Medical Centre, and found a 58% increase in the overall length from 1982 to 1989. In 2004, Sharp et al. reviewed the length of ICDs used in clinical oncology studies in the United States and concluded that the average of approximately 11 pages probably were too long and complex for the patients to read and therefore negatively influenced their understanding of the ICDs [14]. Beardsley et al. studied ICDs from 27 oncological trials carried out across four hospitals in Australia. The number of pages had increased from seven to 11 from 2000 to 2005. Beardsley et al. [15] also compared the content of the ICDs with the National Cancer Institute consent form template and found that the ICDs still missed important information even though they had become longer.
Long ICDs have been discussed and considered problematic in several studies. It has been shown that patients may skim most of the document rather than read it thoroughly [16]. Other studies raise concerns about comprehensive and long ICDs preventing communication of the central aspects of the information since it may be a challenge for the patients to sort out the information that is important to them [15, 17–19]. This may be considered especially important in oncology studies since cancer for most patients is a life-threatening disease. When cancer patients enter studies, most of them are close to disclosure of the primary diagnosis or of tumor progression. In situations like this, the patients are distressed and in a psychological state which probably affect their ability to understand and interpret large amounts of complex information. For this reason, it is recommended that the physician individualizes and portion out detailed information about disease and treatment in a ‘breaking bad news’ situation.
While studies have shown that the length of the ICDs has increased, no previous studies have analyzed what causes the increase in length. The aim of the present study was to investigate whether the length of ICDs for oncological trials in Norway have increased from 1987 to 2007 and to examine what information has been added during this period.
|
methods |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
The selection of ICDs is shown in Figure 1. All clinical trials on cancer patients approved by the REC in the central region of Norway from 1987 to 2007 were identified (n = 253). In 49 trials, written ICDs were missing or not required and these studies were excluded. All ICDs from phase II and III studies testing drugs in cancer patients (e.g. medical treatment of cancer pain, cytotoxic compounds) (n = 41) were included because these studies were considered to have the longest and most complex ICDs. From the remaining 163 trials, ICDs from three projects per year were randomly selected. For years with three or fewer ICDs, all ICDs were included. A total of 87 ICDs were thus selected for the analyses (Figure 1).
|
The ICDs were read into Microsoft Word® format using Scansoft's Omnipage SE® text recognition program. After proofreading, the ICDs were analyzed on number of characters, words, sentences, paragraphs and pages using Microsoft Word 2003® ‘word count’ function.
The ICDs were further analyzed on content in terms of 47 predefined components (Table 1). The list was constructed from the recommendations in The Declaration of Helsinki section 22 that was split up by the authors and found to constitute 13 different components. Thirty-two additional components were identified in the RECs 2004 checklist on what to include in an ICD [10]. Since Norwegian REC base their ICD content guidelines on the recommendations from CIOMS and ICH Harmonised Tripartite Guideline for Good Clinical Practice, these lists were cross-checked [9, 20]. Two additional components [nos 46 and 47 (Table 1)] were identified in the CIOMS 2002 guideline. Components 1–17 were defined as basic components, giving information about fundamental medical and ethical problems, important to the patient in the consent process. Components 18–47 were defined as formal components giving information about formalities.
|
The presence of a component was scored 1, the absence scored 0 and a total score for each ICD was calculated by adding the scores. To ensure reliable scoring, all ICDs from 1987 to 1992 (n = 15) and 2002 to 2007 (n = 32) period were analyzed by two investigators (OB and KS) separately and the results were subsequently compared. The investigators agreed in 88% of the cases. Cases where the investigators did not agree were discussed and an agreed-upon score was obtained. The rest of the ICDs were scored by one of the investigators on the basis of this process.
statistical methods
All material was analyzed using SPSS® 14.0. Year was set as the independent variable in the analyses and number of components and words as the dependent variables. Number of words and components per year were plotted using simple scatter diagram and regression lines added using ‘fit line at total’ function. To better visualize the development in number of words and components per year, regression line was drawn using the Loess function with setting ‘Epanechnikov’ and fit at 50% of points. Loess is a locally weighted polynomial regression [21].
|
results |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
The mean number of words increased significantly from the first 3-year period 1987–1989 [n = 338, standard deviation (SD) = 67] to the last period 2005–2007 (n = 1087, SD = 594) (P < 0.0001) (Table 2, Figure 3). As shown in Table 2, there were variations in the number of words between the ICDs throughout the entire observation period, from a minimum of 165 words to a maximum of 2345 words. The variation in number of words was more pronounced from 1993 to 2007 than from 1987 to 1992 (Table 2, Figure 3).
|
|
The mean number of components increased significantly from the first 3-year period 1987–1989 (n = 9) to the last period 2005–2007 (n = 25) (P < 0.0001) (Table 2). As shown in Figure 2, the most pronounced increase in the total number of components was observed during the last 8–9 years. Also for the components, there was a major variation from a minimum of four to a maximum of 36 components (Table 2).
|
For the components related to basic information, a steady increase was observed throughout the entire observation period from a mean of seven (range 3–11) in 1987–1989 to 14 (range 10–17) in 2005–2007 (Figure 2). The mean number of components about formalities was two in 1987–1989 (range 0–4) and 11 (range 3–20) in 2005–2007. The steep increase in the total number of components during the last 8–9 years was primarily caused by a substantial increase in components involving formalities (Figure 2).
As expected, the total number of words and words regarding formal and basic information is approximately reflected by the number of components (Table 2, Figures 2 and 3). The number of words addressing the components related to basic information increased from 302 (range 184–464) in 1987–1989 to 833 (range 320–1807) in the 2005–2007 period (Table 2, Figure 3). The number of words addressing the components related to formalities increased the most with a seven doubling from 35 words (range 0–92) in 1987–1989 to 254 (26–538) in the 2005–2007 period.
|
discussion |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
Our study showed that the average length of ICDs in Norwegian oncology studies has increased significantly from 1987 to 2007. This is in line with the study by LoVerde et al. from 1989 that found a 58% increase in the overall length in research consent documents from the Denver Veterans Administration Medical Centre from 1982 to 1989 as well as Beardsley et al. who found an average of 4 page increase in 27 oncological trials carried out across four hospitals in Australia from 2000 to 2005 [13, 15]. No previous studies have concluded that some of the increase in length could be explained by an increase in the quantity of text regarding formalities.
Most cancer patients are elderly and sick and probably have limitations in their ability to read and take in large amounts of information in a critical situation at time of inclusion in a study. Many of the older patients also have shorter education which may influence their reading skills. Most studies are proposed to the cancer patients either at the time of primary diagnosis or at the time of tumor progression. At both these situations, it is well documented that patients are distressed and potentially less prepared for processing complex information.
The written ICDs containing all study information are important and necessary because they give the patients a chance to portion the information and read a little at a time. Additionally, the patients have all the information on paper if they need to review it later. Nevertheless most researchers seem to agree that long, extensive ICDs are a potential threat to the validity of the information process. Although few data confirm this, research in other settings have shown that people are unlikely to read entire documents containing >1000 words in an educational context [14, 22]. Increased length may result in patients skimming most of the document rather than reading it thoroughly [16]. Comprehensive and long ICDs could prevent communication of the important and necessary information since it may be a challenge for the patients to sort out the information that is important to them [15, 17–19]. Increased number of components in the ICDs might be problematic in itself. In addition to the increase in length, they increase the complexity of the text introducing more topics. The human short-term memory has a capacity limit of about four items [23]. In that respect, even the short and plain ICDs are challenging enough. Increased number of components and topics in the ICDs might at any rate exceed the patients’ memory capacity and reduce the understanding of the important topics of the ICDs [24, 25].
The process before obtaining the informed consent usually includes a written information part, found in the ICD, as well as an oral communication from the physician. The patients’ trust in their physician has in many studies been shown to be the single most important factor for their decision whether to participate in a study [26]. Long ICDs might challenge this even more because long ICDs could reduce patients’ ability to read and understand the written information. Consequently, patients may be more likely to make a decision on the basis of the influence from their treating physician, rather than their own knowledge about the study.
Taking the researchers’ view, writing good informative ICDs is a challenge. In elucidation of today's claims to the ICDs, it is a demanding task to both satisfy patients’ needs of short concise information and simultaneously fulfilling the recommendations and claims from different authorities.
There could be different reasons for the observed increase in the length of the ICDs. Newer studies might have more complex and advanced procedures than before and need more elaboration as might be reflected in the increase in the amount of text regarding basic information. Text regarding formalities has, however, had the biggest relative increase in length in the time period. There has been a substantial increase in space given to text dealing with formalities such as juridical and financial matters, insurance and data safety and storage. It is not clear why the amount of text regarding formalities increases. However, the RECs in Norway follow the guidelines in the Declaration of Helsinki [6], CIOMS [9] and ICH Harmonised Tripartite Guideline for Good Clinical Practice [20]. CIOMS guidelines have changed considerably from 1982 to 2002. In the guideline to ICDs, one can see increased claims to the content of the ICDs regarding juridical and financial matters, insurance and data safety and storage [7–9].
With increased focus on the individual's autonomy and self-determination, better information to the individual is a natural consequence. This might be one explanation to the increased amount of claims to the content of the ICDs. The ICDs function as a guide for the patients during their study participation as well as researchers’ need to give necessary information on the important aspects of the study should not be underrated. Most claims to the ICDs are basically raised in best intent and to enlighten the patients as much as possible. The legal aspect of medicine is getting even more important. However, one might speculate that it could seem the trend in today's society towards increased number of insurance arrangements has a defensive purpose to hold a retreat open and to protect oneself from lawsuits. This trend might influence the different authorities’ requirements to patient's information to prevent misunderstandings and lawsuits in posterity.
Another possible contribution to the increased length and number of formalities in the ICDs is that studies get more internationalized. This implies that the legalization in one country influences the way the ICDs are written in other countries. International studies prepare master ICDs adapted to the country of origin. These are often used in the same or similar version in the other participating countries and influence the length and content of the ICDs.
The aim of ICDs is to achieve consent from truly informed patients. If increased information results in increased length and number of components, not necessarily improving patients’ understanding, one could question the purpose of constant new claims. To our knowledge, none of the new claims to the ICDs are evidence based. Controlled trials could be carried out to investigate whether there are any benefits of adding more information to the ICDs as well as investigate what type of information is needed in order for a patient to undertake a valid independent decision.
There are some limitations to be commented upon. First of all, the number of projects varied among the three-year periods from 7 in 1990–1992 to 17 in 2005–2007 (mean 12.4 +-SD 4.1). This was due to some variation in the number of approved clinical studies on oncology from the REC and that some studies did not require ICDs or the ICDs were missing. The study did not analyze industry and nonindustry studies separately. Nor did it distinguish between studies being multicentre, of national or international origin. It had to be done this way because most studies were lacking information about these aspects. The number of components was analyzed by reading the ICDs thoroughly. Even though an agreed-upon score were obtained between the authors, there is always a matter of interpretation.
|
conclusions |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
The length of ICDs for clinical oncology studies in Norway has increased substantially from 1987 till 2007. The increase could partly be explained by an increase in the amount of text dealing with juridical and financial matters, insurance and data safety and storage. This development increasingly demands competent readers and might in worst case prohibit truly informed consents. It is time to discuss whether all claims to the content are necessary and try to come up with international guidelines that secure shorter and more readable ICDs for the patients. Studies should be carried out to investigate whether more claims to the content of ICDs are beneficial.
|
funding |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, St Olavs Hospital, N-7005 Trondheim, Norway.
Received for publication May 26, 2008. Accepted for publication August 6, 2008.
|
References |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionconclusionsfundingReferences |
|---|
1. Shuster E. Fifty years later: the significance of the Nuremberg Code. N Engl J Med (1997) 337:1436–1440.
2. Beecher HK, Pappworth M. Informed consent in human experimentation and the physicians response. In: Informed Consent in Medical Research.—Doyal L, Tobias JS, eds. (2001) London: BMJ. 20–27.
3. Norwegian Law on Patients Rights. LOV 1999-07-02 nr 63: Lov om pasientrettigheter (pasientrettighetsloven) (1999).
4. Agard A, Hermeren G, Herlitz J. Patients' experiences of intervention trials on the treatment of myocardial infarction: is it time to adjust the informed consent procedure to the patient's capacity? Heart (2001) 86:632–637.
5. Cox K. Informed consent and decision-making: patients' experiences of the process of recruitment to phases I and II anti-cancer drug trials. Patient Educ Couns (2002) 46:31–38.[CrossRef][Web of Science][Medline]
6. World Medical Association. The Declaration of Helsinki. update 09.10.2004. Available at http://www.wma.net/e/policy/b3.htm (10 March 2008, date last accessed).
7. Bankowski Z, Howard-Jones N. Human experimentation and medical ethics. In: Proceedings of the XV CIOMS Round Table Conference Manila 13–16 September 1981—Bankowski Z, Howard-Jones N, eds. (1982) Geneva: The Council For International Organizations of Medical Sciences.
8. Bankowski Z. International Ethical Guidelines for Biomedical Research Involving Human Subjects (1993) Geneva: CIOMS.
9. Council for International Organizations of Medical Sciences (CIOMS). International Ethical Guidelines for Biomedical Research Involving Human Subjects (2002) Switzerland: CIOMS. http://www.cioms.ch/frame_guidelines_nov_2002.htm (15 November 2002, date last accessed).
10. Regional Committee for Medical and Health Research Ethics. ‘REKs veiledning for utfylling av søknadsskjema’ (Guideline in Norwegian). Available at http://www.etikkom.no (5 May 2004, date last accessed).
11. Norwegian Regulations on Medicine Testing. Forskrift om klinisk utprøving av legemidler til mennesker (2003).
12. Norwegian Biobank Law. 21 December 2002.
13. LoVerde ME, Prochazka AV, Byyny RL. Research consent forms: continued unreadability and increasing length. J Gen Intern Med (1989) 4:410–412.[Web of Science][Medline]
14. Sharp SM. Consent documents for oncology trials: does anybody read these things? Am J Clin Oncol (2004) 27:570–575.[CrossRef][Web of Science][Medline]
15. Beardsley E, Jefford M, Mileshkin L. Longer consent forms for clinical trials compromise patient understanding: so why are they lengthening? J Clin Oncol (2007) 25:e13–e14.
16. Rugege-Hakiza SE, Glynn SA, Hutching ST, et al. Do blood donors read and understand screening educational materials? Transfusion (2003) 43:1075–1083.[Medline]
17. Stanley BM, Walters DJ, Maddern GJ. Informed consent: how much information is enough? Aust N Z J Surg (1998) 68:788–791.[Web of Science][Medline]
18. Davis TC, Holcombe RF, Berkel HJ, et al. Informed consent for clinical trials: a comparative study of standard versus simplified forms. J Natl Cancer Inst (1998) 90:668–674.
19. Sharp SM. The problem of readability of informed consent documents for clinical trials of investigational drugs and devices. United States considerations. Drug Inform J (2004) 38:353–359.[Web of Science]
20. ICH Expert Working Group. ICH Harmonised Tripartite Guideline for Good Clinical Practice. Brookwood: Brookwood Medical Publications Ltd, 1996. Available at http://www.ich.org/LOB/media/MEDIA482.pdf (10 March 2008, date last accessed).
21. Cleveland WS, Devlin SJ. Locally weighted regression: an approach to regression analysis by local fitting. J Am Stat Assoc (1988) 83:596–610.[CrossRef][Web of Science]
22. Brockett RG. Developing written learning material: a proactive approach. Lifelong Learn (1984) 7:16–18.
23. Cowan N. The magical number 4 in short-term memory: a reconsideration of mental storage capacity. Behav Brain Sci (2001) 24:87–114.[CrossRef][Web of Science][Medline]
24. Hochhauser M. Memory overload: the impossibility of informed consent. Appl Clin Trials (Nov.1 2005) Available at http://appliedclinicaltrialsonline.findpharma.com/appliedclinicaltrials/Sites/Memory-Overload-The-Impossibility-of-Informed-Cons/ArticleStandard/Article/detail/192756?searchString=Memory%20overload (10 March 2008, date last accessed).
25. Hochhauser M. Informed consent: reading and understanding are not the same. Appl Clin Trials (April 1. 2004) Available at http://appliedclinicaltrialsonline.findpharma.com/appliedclinicaltrials/article/articleDetail.jsp?id=90594&sk=&date=&&pageID=1 (10 March 2008, date last accessed).
26. Kass NE, Sugarman J, Faden R, Schoch-Spana M. Trust, the fragile foundation of contemporary biomedical research. Hastings Cent Rep (1996) 26:25–29.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||