Participant recruitment strategies in psychosocial oncology research: A comparison of in-person and telephone approaches

Participant recruitment and retention are critical for the completion of clinical research and are an important element of determining a study's validity, interpretability, and generalizability. Research suggests that less than one-third of studies meet recruitment goals [1]. Thus, clinical researchers often struggle to achieve recruitment goals for psychosocial research. This may be especially true in the medical setting due to patient and clinician priorities. To optimize resources in psychosocial research, recruitment approaches resulting in the best outcomes for recruitment, retention, and overall study efficacy should be determined. No research has examined the effect of initial approach type (i.e., in-person or telephone) on both recruitment and retention rates.

Two recruitment strategies were compared: 1) Within one psychotherapy intervention study (i.e., male and female colorectal and anal cancer survivors; Table 1, S1) and 2) Between two separate longitudinal questionnaire-based studies targeting the same population (i.e., adult men with prostate cancer; Table 1, S2 & S3). Questionnaires across studies included similarly sensitive items. Two recruitment strategies were used: 1) In-person recruitment: patients were approached in clinics; a research assistant (RA) explained the study and consented interested patients; 2) Telephone recruitment: patients were sent a letter from the principal investigator and/or clinician introducing the study; a RA then contacted the patient by telephone to explain the study and verbally consent interested patients. All studies were approved by the Memorial Sloan Kettering Cancer Center IRB.

Table 1.

Study summaries.

Study	Participantsa	Purpose	Design	Recruitment method
Study 1 (S1): Sexual health intervention	Male and female anal and rectal cancer survivors	To investigate the efficacy of a sexual health intervention	RCT (4 sessions and 3 questionnaires over 8-months)	Telephone (N = 146) In-person (N = 332)
Study 2 (S2): Decision-making in rising PSA patients	Prostate cancer survivors with a rising PSA	To explore treatment decision-making processes	Longitudinal (5 questionnaires over 1-year)	Telephone (N = 246)
Study 3 (S3): Decision-making in early stage prostate cancer	Early stage prostate cancer patients eligible for active surveillance	To evaluate the impact of anxiety on decision-making	Longitudinal (6 questionnaires over 2-years)	In-person (N = 324)

Note. PSA = Prostate-Specific Antigen.

^aAll studies included participants who were ≥ 18 years old.

The intervention study (S1) used both recruitment strategies, and the effectiveness of these approaches was compared within this study. The two questionnaire-based studies (S2 and S3) included a similar population and set of questionnaires but differed in the recruitment strategy: S2— telephone; S3— in-person.

Three comparisons were evaluated: 1) Recruitment rate— percent of patients recruited out of all approached; 2) Retention rate— percent of patients who complete protocol requirements out of all enrolled; 3) Approach effcacy — percent of participants who completed protocol requirements out of total approached. Chi-square tests were used to determine differences between approach types. Chi-square tests and independent samples t-tests were used to identify potential demographic differences in participants enrolled by telephone versus in-person.

For the intervention study (S1, Table 2), 332 participants were approached in-person and 146 by telephone. In-person recruitment demonstrated a higher recruitment rate compared to telephone recruitment (42% vs. 23%, p < .001, RR = 1.84), while retention rates between the in-person and telephone recruitment were similar (54% vs. 67%, p = .24, RR = 0.80). In-person recruitment demonstrated a higher overall approach efficacy compared to telephone recruitment (22% vs. 15%, p = .08, RR = 1.48).

Table 2.

Recruitment, retention, and overall approach efficacy.

Intervention study comparisons (within S1)
Comparison	Telephone (n = 146)		In-person (n = 332)		χ2	p	RR
	n	%	n	%
Recruitment ratio	33	23	138	42	15.87	< 0.001	1.84
Retention ratio	22	67	74	54	1.84	0.24	0.80
Overall efficacya	22	15	74	22	3.29	0.08	1.48
Questionnaire-based study comparisons (between S2 and S3)
Comparison	Telephone (S2) (n = 246)		In-person (S3) (n = 324)		χ2	p	RR
	n	%	n	%
Recruitment ratio	100	41	224	69	46.25	< 0.001	1.70
Retention ratio	80	80	84	38	49.96	< 0.001	0.47
Overall efficacy	80	33	84	26	2.97	0.08	0.80

^aOverall efficacy = % of participants who completed requirements out of total patients approached.

For the questionnaire-based studies (S2 and S3, Table 2), in-person recruitment (S3) demonstrated a higher recruitment rate compared to telephone recruitment (S2) (69% vs. 41%, p < .001, RR = 1.70); however, in-person recruitment demonstrated a significantly lower retention rate compared to telephone recruitment (S3) (38% vs. 80%, p < .001, RR = 0.47). Thus, in-person recruitment demonstrated lower approach efficacy compared to telephone recruitment (26% vs. 33%, p = .08, RR = 0.80).

There were no significant differences by approach type in race, ethnicity, or gender among those enrolled across studies. In the between-study comparison (S2 vs. S3), those recruited by telephone were significantly older than those recruited in-person (70 vs. 59 years, p < .001).

While in-person approaches were more effective for initial recruitment and telephone approaches were more effective for retention, overall effectiveness of approach differed between study designs. In-person approaches may be more effective for initial recruitment due to an acquiescence bias [2], as individuals may agree to participate for a range of reasons (e.g., diffculty refusing, feeling a sense of “owing” their treating institution).

In-person approaches appeared better for overall approach efficacy in the psychotherapy intervention study. Intervention studies are a larger time commitment compared to questionnaire-based studies, and accordingly, individuals may be more likely to consent only if they genuinely perceive a need for participation. This intervention study addressed sexual health and it could be that for sensitive topics, a warm approach by trained staff eases patient discomfort compared to receiving a letter.

Race, ethnicity, and gender were not differentially represented in the telephone or in-person approach groups. Thus, the results were not confounded by demographic differences among the groups. However, this study can only report characteristics of enrolled participants, not those who declined participation. It is possible there would be demographic differences observed among the total number of patients approached rather than just those who enrolled. Interestingly, those enrolled via telephone were significantly older than those enrolled in-person (i.e., S2 and S3). This suggests that older cancer patients (in this case, older men) may prefer a telephone approach to an in-person approach as it may be perceived as more trustworthy, with more time to review the material and make an informed decision.

Conversely, the overall approach efficacy for the questionnaire-based study was higher among participants with telephone approaches. Patients may recognize questionnaires as a smaller time commitment, resulting in increased initial recruitment across both approach types. However, perhaps due to the acquiescence bias, participants may not ultimately be genuinely interested in participation, and retention rates are consequently much lower for in-person approaches. The higher attrition in the in-person group suggests that the significant time and resources devoted to in-person recruitment may not optimize resources; instead, targeted invitations for participation by mail with telephone follow-up could provide the most sustainable yield.

In sum, differences in effectiveness may be attributed to varied patient needs, social obligation, and perceived time commitment. Identifying the most effective recruitment approach for different types of psychosocial research can inform the design of future studies, and successful recruitment and retention practices have the potential to optimize study pace and completion across settings.