|Year : 2012 | Volume
| Issue : 6 | Page : 94-99
Prospective analysis of reasons for non-enrollment in a phase III randomized controlled trial
Vedang Murthy1, Kasturi R Awatagiri1, Pramod K Tike1, Sarbani Ghosh-Laskar2, Tejpal Gupta1, Ashwini Budrukkar2, Mandar S Deshpande2, Devendra A Chaukar2, Gouri H Pantavaidya2, Jai Prakash Agarwal2
1 Department of Radiation Oncology, Epidemiology and Clinical Trial Unit, Advanced Centre for Treatment, Research and Education in Cancer, Kharghar, Navi Mumbai, Maharashtra, India
2 Department of Radiation and Surgical Oncology, Tata Memorial Hospital, Mumbai, Maharashtra, India
|Date of Web Publication||24-Jan-2012|
Department of Radiation Oncology, Tata Memorial Centre, Advanced Centre for Treatment, Research and Education in Cancer, Kharghar, Sector 22. Navi Mumbai - 410210, Maharashtra
Source of Support: None, Conflict of Interest: None
Aim: This study aims to provide information on the accrual rate and to identify the reasons for non-enrollment of oral cancer patients into a clinical trial.
Setting and Design: Prospective study conducted at the Tertiary Cancer Centre (India).
Materials and Methods: Patients eligible and screened for the oral cancer adjuvant therapy (OCAT) were logged prospectively and reasons for non-enrollment were documented which were broadly divided into patient and trial related.
Statistical Analysis Used: Demographic characteristics of the non-enrolees were compared with the enrolled. Factors predicting non-enrollment were analyzed using multivariate logistic regression test.
Results: A total of 1335 patients with locally advanced cancer of the oral cavity were screened of whom 498 (37%) could be enrolled. Among non enrolled 837 patients, 182 (22%) had the trial-related reasons and 655 (78%) had patient-related reasons. Most important patient-related reasons were patients' preference of taking treatment closer to their native place (26.2%), lack of interest (16.8%) in trial participation. Anticipated poor compliance to treatment (5.9%) and follow-up (6.6%), inability to start treatment in time (6.2%) were important trial-related reasons for non-enrollment. Multivariate analysis identified the genders (female), education (illiterate), occupation (laborer) and non availability of support system in the city as significant predictors of non-enrollment.
Conclusions: Both trial design and patient factors were important causes of non enrollment in eligible patients. Patients' need for being closer to home and refusal to participate were the most common reasons for non-enrollment.
Keywords: Non enrollment, oral cancer, randomized control trial, reasons, trial design
|How to cite this article:|
Murthy V, Awatagiri KR, Tike PK, Ghosh-Laskar S, Gupta T, Budrukkar A, Deshpande MS, Chaukar DA, Pantavaidya GH, Agarwal JP. Prospective analysis of reasons for non-enrollment in a phase III randomized controlled trial. J Can Res Ther 2012;8, Suppl S2:94-9
|How to cite this URL:|
Murthy V, Awatagiri KR, Tike PK, Ghosh-Laskar S, Gupta T, Budrukkar A, Deshpande MS, Chaukar DA, Pantavaidya GH, Agarwal JP. Prospective analysis of reasons for non-enrollment in a phase III randomized controlled trial. J Can Res Ther [serial online] 2012 [cited 2020 Feb 26];8:94-9. Available from: http://www.cancerjournal.net/text.asp?2012/8/6/94/92221
| > Introduction|| |
Clinical trials are crucial in improving cancer treatment. Participation in trials is often associated with a higher survival rate  and offers patients newer and possibly better treatment options.  However, several restrictions within the process of conducting a trial of which non-enrollment is a significant barrier for conducting clinical trials leading to delay in completion of the study.  Various studies from west have reported accrual barriers in trials, ,, but application of these results to Indian population is difficult given the socio-economic and cultural variations. No data is available from India regarding causes of non-enrollment in clinical trials.
| > Materials and Methods|| |
We have prospectively identified the reasons for non-enrollment within the ongoing oral cancer adjuvant therapy (OCAT) trial conducted at the Tertiary Cancer Centre (India). In this study, all patients after undergoing surgery for advanced oral cancers are randomized into three arms [Figure 1]. Patients in the standard arm are given conventional post-operative radiotherapy (five days per week) while the patients in study arms are given concurrent chemo radiotherapy and accelerated radiotherapy (six days per week). All patients are expected to be followed up for at least five years after treatment for evidence of recurrence and documenting toxicity. Patients are recruited from all over India, who in the opinion of the investigators have understood the protocol, its implications and are willing to complete therapy and follow-up. Patients are screened and counseled regarding the study by the investigators and their baseline data collected. Based on their eligibility, patients are enrolled in the trial and the non-enrolled patients are offered treatment according to the standard practice. A detailed log is maintained for the enrollment status and non-enrollment rates from the beginning of the study. The OCAT trial was approved by the Institutional Review Board (IRB) and started accrual in June 2005 and this analysis is of patients screened for the study up to May 2010.
Statistical analysis was performed using SPSS 15 software. Descriptive statistics on patient characteristics were generated. Associations were assessed using the cross tabulation and statistical significance by the Chi-squared test and a two-sided P value was set less than 0.05. Multivariate logistic regression was done using clinical trial enrollment versus non-enrollment as the dependant variable (yes vs no) and factors such as gender, education, residence with city, occupation, marital status, financial dependence and age as predictors for all the patients.
| > Results|| |
A total of 1335 patients were screened for the trial. Demographic details of the entire study population (enrolled and non-enrollees) are shown in [Table 1]. The mean age was 47.4 years; and male dominance 79.5% was seen as expected. Nearly a quarter of the patients were illiterate and a majority had completed some form of education. It was observed that half of the screened subjects were residents outside Maharashtra state. A majority patients were Hindu 85.4% (n =1136) and 90.8% (n =1212) were married. A majority of the patients screened (82.4%) did not have local support in the form of residence, relatives and friends in Mumbai. Nearly a third (n = 407) were laborers while 21% (n= 280) were financially dependent. In terms of billing for hospital service, only 13.2% patients were in private category.
The study showed that non-enrollment rates of 63% (n = 837) outnumbered the enrollment rate at 37% (n = 498). These non-enrolled patients were eligible per say, but on a detailed history and examination found to be unsuitable for trial due to various reasons. The reasons for non-enrollment were broadly divided into two categories i.e. trial-related and patient-related reasons for non-enrollment. There were 182 non-enrollments as a result of trial-related reasons [Table 2]. The most common factors were unreliability for regular (telephonic contact or hospital visit) follow-up accounting for 6.6% and poor compliance to treatment as assessed by the investigators in 5.9% of the patients. Trial design mandated beginning the adjuvant treatment within 56 days of surgery and 6.2% patients became ineligible due to delayed wound healing or post-operative complications and reported late to the trial unit. Disease progression and detection of abnormalities in the hematological and renal parameter were observed in 3.1% of non-enrollees.
[Table 2] describes the various patient-related factors which hindered their participation in the trial. Patients' refusal (no explanation provided or asked) accounted for 16.8%. Just over a quarter (26.2%) of the patients preferred to take adjuvant treatment at other hospitals or at their native place after surgery. Ninety nine (11.8%) patients were lost to follow-up after surgery and did not turn up for enrollment or treatment. A few (1.2%) patients were concerned about side-effects or toxicities including fear of injections and patients had low comprehension or had a language problem making it difficult for them to understand the trial and its processes.
[Table 3] summarizes the univariate analysis of the factors affecting enrollment. Age, gender, literacy, support system in the city, financial dependence, occupation and marital status were significant while there was no significant difference between enrollees and non-enrollees with respect to patients' residence in the state and religion.
On multivariate analysis [Table 4], the significant predictors of non-enrollment were gender (female), education status (illiterate), occupation (laborers) and non availability of support system in the city. Occupation, marital status, financial dependence and age did not remain significant predictors of non-enrollment.
|Table 4: Multivariate logistic regression model of predictors of non enrollment in OCAT trial|
Click here to view
| > Discussion|| |
This is the first prospective study to analyze and report in detail the issue of reasons for non-enrollment in the setting of a phase III trial in India. After screening 1335 potential participants, 837 patients were non-enrolled due to various trial- and patient-related reasons. Only 498 of the 1335 screened patients were enrolled in the study which resulted in an accrual rate of 37%.
Published literature from Europe and North America shows an accrual rate ranging from 10% to 51% ,,,, and the reasons for non enrollment are summarized in [Table 5]. However, it should be observed that the majority of patients excluded from these trials showed patient characteristics and disease characteristics that were completely different from the present study reflecting the socio-demographic characteristics of the population participated in the study. The common reasons for non-enrollment cited in the western literature are insurance denials, strict and narrow eligibility criteria, patient's refusal to be treated in an experiment, lack of availability of trials and the concern of randomization as expressed by the screened patients. In the current study, the factors were different and are discussed below.
|Table 5: Summary of published data on non enrollment from U.S.A and Europe|
Click here to view
Follow-up plays a critical role in the success of a trial and good compliance to the scheduled follow up leads to higher enrollment. Although difficult to quantify, compliance to follow up is frequently associated with the attitude and behavioral dynamics of physicians and participants.  In the present protocol, the study period was planned for five years with a minimum follow-up for two years. It was observed that only 6.6% participants were deemed as unlikely to follow-up as per the requirement of the study protocol. The low rate was possibly because the protocol design mandated the follow up would be as per routine clinical practice and there were no extra scheduled visits. The patients deemed unlikely to follow up lacked even basic contact details such as telephone numbers and residential address making them potentially untraceable in case if they did not visit a clinician on the scheduled follow-up date.
Patients in this study had undergone major surgery and were scheduled to receive adjuvant therapy with radiotherapy and chemotherapy. Non enrollment due to poor performance status was primarily because of safety issues, and the patients' inability to receive chemotherapy if randomized to the chemoradiation arm. Previous studies reported that poor performance status of the trial participant is the most common reason for not recruiting patients in a clinical trial. The majority of the trials are designed for patients with good performance status. , In the current protocol, patients were recruited with the Kernofsky performance status (KPS) of ≥70 making 49 patients (5.9%) non-enrolled. It must be noted that some subjectivity would always be present in determining the KPS and hence the eligibility in a trial especially when it is conducted on a large scale in a busy institution.
Overall treatment time has been reported to be a significant predictor of survival in patients undergoing treatment for head and neck cancers.  The OCAT protocol specified that the adjuvant treatment should start within 56 days of primary surgery. Non enrollment due to this specification in the protocol made 52 patients ineligible (6.2%). This included patients with post operative wound complications and those who reported late at the trial unit due to various reasons (mainly social and financial).
A total of 78% (655/837) patients were otherwise eligible, but refused to participate citing various reasons like financial burden, geographical barrier, social issues, personal and language problems and concern about side-effects of the treatment (reported here as patient related factors).
The most common reason (26.2%) for non enrollment seen in this study is that patients wanted to take treatment closer to their homes around the country. Over half of our patients belonged to a state outside Maharashtra as well as lacked social support in the form of residence in Mumbai or friends or relatives to accompany them for treatment. As the duration of the study intervention was up to six to eight weeks, relatives found it difficult to accompany the patients throughout their treatment. The social issues that influenced the decision to participate include time constraints, interferences with work and domestic responsibilities. Residence of the patients, lack of commuting facilities and access to the hospital were found to be troublesome for several potential participants. Transportation as a barrier to clinical trial participation has been reported by Baggastrom et al.  in patients with non small cell lung cancer. The most common reasons for non enrollment they reported in spite of eligibility were patient refusal (8.7%) and problems with transportation and distance from the medical center (7.1%) in Washington. The higher rate seen in our study is expected given the vast difference in the transportation infrastructure between two cities where the studies are reported from.
Another substantial reason (16.8%) for non enrollment in our study was that the patients simply declined to participate without providing any explanation and did not show any interest in trial participation. This was after they were explained in detail about the study. Perquine et al. identified the difficulties in recruitment for a randomized controlled trial involving hystero-salpingography. They reported that the most common reasons for non-enrollment were patient refusal and indecision to take part in the study.  We also had 11.8% patients who did not report for randomization after being served the informed consent form and discussion about the study. Although we were unable to capture the reason for this, it may be related to a number of issues involving their social, financial or medical circumstances and the fact that a clinical trial is a novel concept for most Indian patients.
Financial burden of treatment and need for prolonged stay in an alien city was a major problem for many patients. Although substantial financial support was provided for treatment, trial-related procedures like feeding tubes (endonasal and gastric) and for chemotherapy drug, some patients could not even afford daily expenses. While the non enrollment rate due to a stated financial reason was 7.6%, but it would be safe to assume that a much larger number of enrolled patients also had major financial issues which would affect treatment and follow up compliance as well. Socioeconomic circumstance has been reported as a barrier to clinical trial enrollment even in the western literature. Satreen et al. examined the impact of socio-economic factors on accrual to National Cancer Institute sponsored clinical trials. They reported high accrual rate of participants from geographical areas with higher socio-economic status.  Also, Trimbal et al. studied the underrepresentation of the elderly population in suitable cancer treatment trials and reported that a low accrual rate has been observed in poorer socio-economic class and also in elderly people who might lack social, logistical and monetary assistance.  We were not able to particularly look at the elderly population as the eligibility for trial was < 65 years of age.
Defaulters who did not report to trial unit after screening formed a substantial part of non-enrolled patients (11.8%). Sometimes the negative attitude and false perception of family members can influence a patient's decision and relationship with treating doctors making them not to participate in a clinical trial.  Informed consent process plays a very important role in accrual of patients. Despite initial misgiving, we found that only ten patients could not comprehend the concept of trial and preferred standard treatment. These patients had difficulty in comprehension of the concept of a clinical trial. In addition, being a tertiary referral center, it serves patients from all over the country and language barrier can be a significant hurdle in consenting for trial. Although as per the IRB mandate, the consenting process was officially conducted in three languages (Hindi, Marathi and English), it was difficult to cater to other regional languages.
The limitation of this study is that it analyzes patients from a single trial that was conducted at a single center which may not be representative of the entire population especially cancers that represent different socio-demographic patterns e.g. breast or prostate cancer. The reasons for non-enrollment might vary in other places within the country and with other cancers. Health insurance data is not available with us to study its impact on non enrollment and compare it with the previously published data.
| > Conclusion|| |
The significant difference between enrolled and non-enrolled patient characteristics was found in the studied population. Both trial design and patient factors were important contributors of non enrollment in eligible patients. Patients' need for being closer to home and their refusal to participate or report to the trial unit were the most common reasons for non enrollment. Corrective measures against the discussed causes should help to improve accrual in future trials. This study may help to benchmark the non-enrollment rates within the context of a large cancer center in the Indian subcontinent so that the trial sample size and end points can be better planned in future studies.
| > Acknowledgements|| |
The author wants to acknowledge Ms Sadhana Kannan for guidance of statistical analysis and Sister Remya Rajesh for valuable help in writing the article.
| > References|| |
|1.||Stiller CA. Centralised treatment, entry to trials and survival. Br J Cancer 1994;70:352-62. |
|2.||Cobau CD. Clinical trials in the community. The community clinical oncology program experience. Cancer 1994;74 (suppl 9):2694-700. |
|3.||Haidich AB, Ioannidis JP. Effect of early patient enrollment on the time to completion and publication of randomized controlled trials. Am J Epidemiol 2001;154:873-80. |
|4.||Lara PN Jr, Higdon R, Lim N, Kwan K, Tanaka M, Lau Derick HM, et al. Prospective evaluation of cancer clinical trial accrual patterns: Iidentifying potential barriers to enrollment. J Clin Oncol 2001;19:1728-33. |
|5.||Sateren WB, Trimble EL, Abrams J, Brawley O, Breen N, Ford L, et al. How Sociodemographics, Presence of Oncology Specialists, and Hospital Cancer Programs Affect Accrual to Cancer Treatment Trials. J Clin Oncol 2002;20:2109-17. |
|6.||Paterniti, DA, Chen MS, Chiechi C, Beckett LA, Horan N, Turrell C, et al. Asian Americans and cancer clinical trials: A mixed-methods approach to understanding awareness and experience. Cancer 2005;104:3015-24. |
|7.||Du W, Gadgeel SM, Simon MS. Predictors of enrollment in lung cancer clinical trials. Cancer 2006;106:420-5. |
|8.||Harter P, Bois A, Schade- Brittinger C, Burges A, Wollschlaeger K, Gropp M, et al. Non-enrolment of ovarian cancer patients in clinical trials: Reasons and background. Ann Oncol 2005;16:1801-5. |
|9.||Baggstrom MQ, Waqar SN, Sezhiyan AK, Gilstrap E, Gao F, Morgensztern D, et al. Barriers to enrollment in non-small cell lung cancer therapeutic clinical trial. J Thorac Oncol 2011;6:98-102. |
|10.||Corrie P, Shaw J, Harris R. Rate limiting factors in recruitment of patients to clinical trials in cancer research: Descriptive study. BMJ 2003;327:320-1. |
|11.||Tangrea JA. Patient participation and compliance in cancer chemoprevention trials: Issues and concerns. Proc Soc Exp Biol Med 1997;216:260-5. |
|12.||Rosenthal DI, Liu L, Lee JH, Vapiwala N, Challan AA, Weinstein GS, et al. Importance of the treatment package time in surgery and postoperative radiation therapy for squamous carcinoma of the head and neck. Head Neck 2002;24:115-26. |
|13.||Perquin DA, de Craen AJ, Helmerhorst FM. Difficulties in recruitment for a randomized controlled trial involving hysterosalpingography. Reprod Health 2006;3:5. |
|14.||Trimble EL, Carter CL, Cain D, Freidlin B, Ungerleider RS, Friedman MA. Representation of older patients in cancer treatment trials. Cancer 1994;74:2208-14. |
|15.||Castel P, Negrier S, Boissel JP. Why don't cancer patients enter clinical trials? A review. Eur J Cancer 2006;42:1744-8. |
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]