The database searching and the forward and backward citation checking yielded 13,278 and 13 papers respectively (13,291 total). 5,008 duplicates were removed resulting in 8,283 articles available for screening (Figure 1). Of these, 7,243 studies did not meet the inclusion criteria based on titles and abstract screening and resulted in 1,040 full-text studies selected for further screening (Figure 1). A total of 1,022 studies were then excluded with 214 having no male specific data, 140 conference abstracts, 115 HIV related research, 106 cancer related research, 71 studies had no included data on barriers or facilitators, 70 studies with a focus on men > 60 years, 69 studies from racial or ethnic minority, 48 Alzheimer’s or dementia research, 46 studies were unrelated to health recruitment and retention, 35 related to illegal drugs, 25 papers were studies with less than 3 study visits, 23 papers were men < 16 years of age, 21 systematic review/review papers, 17 focused on socioeconomically disadvantaged populations, 11 uncompleted studies/study protocol, 9 studies were <12 weeks duration, and 2 fathers in early childhood interventions (Figure 1).
A total of 18 articles remained and data was extracted and included in this review. The oldest of these studies was published from 1976 [40] and the most recent 2021 [41] All of the included studies were conducted in Western countries excepting Cheraghi et al., which was based in Middle East [41]; two were located in United Kingdom [42, 43], two in France [44, 45], one in Finland [46], one in Sweden [47], one in The Netherlands [48], one study across combined European nations [49], seven in North America [40, 50-55] and two in Australia [56, 57] and are described in Table 1. Participant characteristics varied with study focus including participants with specific health conditions, such as overweight [43, 56], having an occupational injury [52], having visited a sexually transmitted infection clinic [48], or being treated for a psychological disorder [46, 50, 53], or habits such as alcohol abuse and smoking [55]. Some studies recruited participants from specific subgroups, including veterans [40], workers of an electricity company [44] and people that had attended a spouse abuse abatement program [50]. All eighteen included studies met the inclusion criteria for age [41-46, 48, 49, 53, 54, 56]. One of the studies was a family cohort study that recruited families of children with cystic fibrosis and congenital heart disease and required participation of both parents [51].
Of the included studies, 14 had male and female participants [41, 42, 44-48, 51-57], with approximately one-third having a predominantly male sample [44, 52, 53, 57]. Four studies recruited only male participants [40, 43, 49, 50] (Table 2). The included studies with mixed gender either described male and female characteristics separately or clearly stated that there were no significant differences in recruitment and retention based on gender. All included studies used a minimum of three study visits or data collection, and the maximum number of study visits or data collections was 95 visits [43]. The minimum study length of included studies was 16 weeks [50] and the maximum study duration was 43 years [47]. All included studies collected demographic data [40-57].
Recruitment
Overall, 12 studies provided information on recruitment rates [42-46, 48-50, 53, 55-57] and 12 provided information on retention rates [40, 43, 44, 46-48, 50-52, 54-56], whereas only 7 studies had information about both recruitment and retention rates [43, 44, 46, 48, 50, 55, 56] (Table 3). A variety of methods for male participant recruitment included advertising [40, 45, 56], letters of invitation [42-44, 49, 52, 55, 57], selection of participants from larger cohort [44, 45, 50, 53], or recruitment from hospitals or registers [43, 46-48, 50, 51, 53]. The most common method was sending letters of invitation, used in 8 out of the 18 studies, and yielded recruitment rates between 16.7% - 76%. Irvine et al., recruited participants through letter of invitation and time space sampling, and reported that time space sampling was difficult, time consuming and only yielded one participant per 11 field visits [43]. Snow et al., used multiple methods for recruitment, including recruitment from work sites and public sites, mass mailing, telephone, media, and referral methods and reported that mass mailing was the best method of these [54]. Rose et al., attributed their high recruitment rates to advertising and therefore people that agreed to participate had done so voluntarily and were more likely to be interested in the study and health interventions in general [40]. To maximise male participation, vanWees et al., adapted their recruitment methods to target male participants by raising awareness and a greater sense of responsibility in terms of male health through flyers or personalised invitations [48].
Barriers
A variety of factors were identified that interfered with male participation in longitudinal research are shown in Table 4. Some of these were situational and included participant death or relocation [40, 41, 44, 47, 49, 51, 53, 54, 56].
Facilitators
Many studies employed a variety of strategies to increase participation for males (Table 5). These varied from offering free medical screening, reminders for appointments, incentives or enrolment of wives to assist in retention. Several studies used a range of strategies, particularly [45, 55, 56], with varying degrees of success.
Quality check
Assessment using the MMAT quality assessment tool determined a high-quality study had an answer of ‘yes’ to each of the MMAT questions, a medium quality study had one ‘no’ or ‘can’t tell’ response, and a poor quality study had two or more ‘no’ or ‘can’t tell’ responses [39] (Table 6). Of the included studies the majority were medium quality (7 out of 18 studies) [40, 42, 44, 45, 48, 51, 53]. These studies generally failed to report blinding of assessors, if confounders were accounted for, or any bias related to non-response within their publications. Seven included studies were of high quality [41, 43, 46, 47, 49, 50, 57], the remaining four were considered to be of poor quality [52, 54-56].
Table 1. General study characteristics and male participation. Studies are listed alphabetically
Author / Year
|
Study Name (if identified)
|
City (or State/ Province/ Region), Country
|
Study Aim/s
|
Study Design
|
Clinical field
|
Type of data collected
|
Cheraghi et al., 2021 [41]
|
Tehran Lipid and Glucose Study (TLGS) [58]
|
Tehran,
Iran
|
Clarify factors associated with non-participation in the TGLS cohort.
Identify sub-groups who are likely to refuse participation and improve retention strategies for these groups
|
Longitudinal cohort study
|
Diabetes and non-communicable diseases (NCDs)
|
Protocol [58] and this paper include data on prevalence of NCDs, risk factors for NCDs, intervention data for intervention undertaken by specific target groups of cohort including schoolchildren, women, high-risk groups. Interventions focused on lifestyle modifications; diet, smoking and physical activity.
|
Crichton et al., 2012 [56]
|
-
|
Adelaide, Australia
|
1. Evaluate the recruitment process, retention of participants and challenges faced in a dairy intervention trial,
2. Provide strategies to combat the difficulties of running long-term dietary intervention trials
|
Randomised controlled dietary intervention
|
Obesity
|
Height, weight, blood pressure, health and dietary questionnaires, anthropometry, and blood samples taken at each visit (cardiometabolic and biochemical assessment), arterial compliance, cognitive assessment and mental health assessment.
|
Goldberg et al., 2006 [44]
|
GAZ and ELectricité (GAZEL) Cohort Study
|
Saint- Maurice, France
|
Determine the socioeconomic, lifestyle, and health factors associated with response to annual mail questionnaires
|
Longitudinal cohort study
|
-
|
Questionnaires on health status, lifestyle, socio-economic and occupational factors.
|
Green et al 2018 [42]
|
Cambridge Centre for Ageing and Neuroscience (Cam-CAN) study [59]
|
Cambridge, United Kingdom
|
Determine the factors that affect response in epidemiological studies
|
Contemporary cross-sectional population-based cohort study
|
Epidemiology and Aging
|
The included study and its accompanying protocol paper [59] collect interview data related to recruitment and inclusion criteria, core cognitive neuroscience (MRI, MEG, cognitive testing), blood pressure, salivary sample and 3 fMRI.
|
Hamberger et al., 2000 [50]
|
-
|
Wisconsin, United States of America SA/ Wisconsin
|
Identify reasons for participant dropout from a spouse abuse abatement program
|
Cohort study
|
Psychology
|
Race, education, age, marital status, employment status, alcohol abuse, referral status, police violence contact, drug activity, acting out, history of child abuse, witnessing violence, violence duration, violence level report
|
Irvine et al., 2017 [43]
|
-
|
Dundee, United Kingdom
|
Assess feasibility of trial for reduction of alcohol consumption in obese men
|
Community-based Intervention
|
Obesity
|
Alcohol consumption measures and body mass index
|
Janus et al., 1997 [51]
|
-
|
Toronto, Canada
|
Determine factors influencing family participation in a longitudinal study
|
Longitudinal birth cohort study
|
Cystic fibrosis, Congenital heart disease
|
Severity of chronic illness, diary completed by parents, developmental level, child temperament, family environment, Mother-Child relationship
|
Kannisto et al., 2017 [46]
|
Mobile.Net
|
Turku, Finland
|
Determine the dropout predictors from a mHealth-based trial and evaluate the effects of tailored short message service (SMS) text message constructed to encourage patient adherence
|
mHealth-Based Randomized Controlled Trial
|
Psychology
|
Participant's quality of life (Q-LES-Q) and satisfaction with the treatment (CSQ-8)
|
Kelfve et al., 2017 [47]
|
Swedish level-of-living
survey (LNU)/ The Swedish panel study of
living conditions of the oldest old (SWEOLD)
|
Stockholm, Sweden
|
Determine how selective survey participation affects the sample
composition, in addition to selective mortality of participants
|
Longitudinal cohort survey study
|
Aging
|
Both LNU and SWEOLD primarily use face-to-face interviews to gather data. The questionnaires used cover a broad range of topics, such as living conditions, family situation, health, health behaviours, and financial resources
|
Lee et al., 2009 [49]
|
The European Male Ageing Study (EMAS)
|
European cities: Florence (Italy), Leuven (Belgium), Lodz (Poland), Malmo ̈ (Sweden), Manchester (United Kingdom), Santiago de Compostela (Spain), Szeged (Hungary) and Tartu (Estonia)
|
Examine aspects of aging in men
|
Prospective cohort study
|
Aging
|
Questionnaire: assess quality of life, depressive symptoms, Adverse Life Events Scale, Physical Activity Scale, International Prostate Symptom Score, previous surgical procedures, sexual function questionnaire. Screening: Physical and cognitive performance, Anthropometry, Calcaneal ultrasound, Food diary, blood samples
|
Markanday et al., 2013 [57]
|
Geelong Osteoporosis Study
|
Geelong, Australia
|
Investigate sex-differences in non-participation at
baseline of the Geelong Osteoporosis Study (GOS)
|
Prospective cohort study
|
Osteoporosis
|
Reasons for not participating
|
Méjean et al., 2014 [45]
|
NutriNet-Santé Study
|
Paris, France
|
Evaluate relationships between participation motives and sociodemographic, health, and lifestyle characteristics of participants in cohort designed to identify nutritional risk or protective factors
for chronic diseases
|
Web-based, Prospective
Cohort study
|
Nutrition
|
Questionnaires assessing dietary intake, physical activity, anthropometrics, lifestyle, and socioeconomic conditions along with health status.
|
Oleske et al., 2007 [52]
|
-
|
3 Midwestern states (include names if available), United States of America
|
Determine influence of demographics, health, and job factors in continued participation of employed persons in a longitudinal intervention study for work-related low-back disorders
|
Randomized clinical trial
|
Work-related low-back disorders
|
Self-reported measures: Back pain frequency and degree of bothersomeness, back pain disability, physical health, mental health, neurogenic symptoms, psychological job strain / Measures: height, weight, body fat percentage, waist and hip circumference, body mass index
|
Olmos-Ochoa et al., 2019 [53]
|
-
|
Los Angeles, United States of America
|
Identify barriers to study participation and retention, in two modes of intervention for serious mental illness
|
Cohort study
|
Psychology
|
Barriers to physical activity and healthy eating
|
Rose et al., 1976 [40]
|
Normative Aging Study of the Veterans Administration
|
Boston, United States of America
|
Determine the non-pathological aspects of aging
|
Longitudinal health cohort
|
Aging
|
Age, social class, geographic stability and health
|
Snow et al., 2007 [54]
|
The Lung Health Study (LHS)
|
Minneapolis, United States of America and Winnipeg, Canada
|
Examine the impact of smoking
cessation coupled with the use of an inhaled bronchodilator on chronic obstructive pulmonary disease
|
Randomized, controlled clinical
trial
|
Chronic obstructive pulmonary disease
|
Demographic variables, alcohol intake, body mass index (BMI), smoking-related variables, past and present illness, lung function, and social support variables
|
Ullman
et al., 1998 [55]
|
Newcomb [60]
|
Los Angeles, United States of America
|
Develop models that differentiate eager, reluctant, and nonresponding participants using participants’ demographics, personality, and drug use characteristics.
|
Longitudinal cohort study
|
Psychosocial health and Substance use
|
The included paper and its accompanying protocol [60] collect data on use of monetised incentives for longitudinal recruitment, participant demographics, drug and alcohol intake, personality and attitudinal traits, including support of science/medicine and social conformity.
|
vanWees et al., 2019 [48]
|
Mathematical models incorporating Psychological determinants: control of Chlamydia Transmission (iMPaCT)
|
Amsterdam, Hollands Noorden, Kennemerland, and Twente, the Netherlands
|
Identify predictors of non-response and loss to follow-up in longitudinal sexual health study
|
Longitudinal cohort study
|
Chlamydia Transmission, Sexually transmitted infections (STI)
|
Data on sexual behaviour, psychological determinants and chlamydia
infections. Participants were tested for chlamydia using nucleic acid amplification tests at enrolment at the STI clinic and through a self-sampling kit sent to a laboratory at six-month follow-up
|
Table 2. Study duration, number of study visits, percentage of male participants of the study, and details of recruitment and retention numbers (n.d indicates that no data was available in the published literature)
Author / Year
|
Study duration
|
Study status at time of publication (Rec Recruitment, Ret Retention, F Follow-up)
|
Number of Study visits/ contacts
|
% of male participants in the study
|
Total number of male participants recruited
|
Total number of male participants retained
|
% male participants retained
|
Cheraghi et al., 2021
|
20 years
|
Ret & F
|
5
|
42.39%
|
n= 4,395
|
n.d#
|
55% >60 years
67% 40-59 years
57% 20-39 years
|
Crichton et al., 2012
|
1 year
|
Rec, Ret & F
|
3
|
27%
|
n= 20; 18-71 years
|
n= 10
|
50%
|
Goldberg et al., 2006
|
11 years
|
F
|
11
|
73%
|
n= 8,550; 40-45 years
n= 6,277; 45-50 years
|
n.d*
|
n.d
|
Green et al., 2018
|
Study launched in 2010a
|
Rec
|
3b
|
43.7%
|
n= 3,315
|
n= 1,172
|
35.4%#
|
Hamberger et al., 2000
|
Recruitment duration: 6 years 2 months, Study duration: 16 weeks
|
Rec, Ret & F
|
16
|
100%
|
n= 534
|
n= 150; 25-34 years
n= 49; < 25 years
n= 76; > 35 years
|
28% 25-34 years
9.2% < 25 years
14.2% 35 years
|
Irvine et al., 2017
|
5 months
|
Rec, Ret & F
|
96
|
100%
|
n= 69
|
n= 59
|
85.5%
|
Janus et al 1997
|
4 years
|
Ret
|
4
|
50%
|
n= 209
|
n= 135
|
64.6%
|
Kannisto et al., 2017
|
1 year
|
Rec, Ret & F
|
24-300c
|
49.17%
|
n= 560
|
n= 227
|
41.0%d
|
Kelfve et al., 2017
|
43 years
|
F
|
7
|
15.2%
|
n= 172
|
n= 134e
|
41.4% 77-87 years
|
Lee et al., 2009
|
6 years
|
Rec
|
3
|
100%
|
n= 3,963
|
n= 3,369
|
38.7% 40-49 years
45.0% 50-59 years
43.2% 60-69 years
34.4% >70 years
|
Markanday et al.,2013
|
10 years
|
Rec
|
3f
|
50.8%
|
n= 2,296
|
n= 1,540
|
67.1%
|
Méjean et al., 2014
|
10 years
|
Rec & Ret
|
11-120g
|
24.1%
|
n= 3,929
|
n= 1,531
|
24.1%
|
Oleske et al., 2007
|
12 months
|
F
|
5
|
79.3%
|
n= 360
|
n= 168
|
83.6% 47.1 + 6.7 years
|
Olmos-Ochoa et al., 2019
|
6 months
|
Rec, Ret & F
|
24 or 30h
|
Overall Study: 93.9%
Sub Study: 83.3%
|
n= 260
n= 40
|
n.di
|
n.dj
|
Rose et al., 1976
|
12 years
|
Rec & Ret
|
n.dj
|
100%
|
n= 2,280
|
n= 2,028
|
88.9%
|
Snow et al., 2007
|
11 years
|
Ret & F
|
6
|
62.4%
|
n= 3,327
|
n= 2,749
|
82.6%
|
Ullman et al., 1998
|
16 years
|
Rec & Ret
|
4
|
37.6%
|
n= 207
|
n= 201
|
97.1%
|
vanWees et al., 2019
|
2 years
|
F
|
4
|
19%
|
n= 324
|
n= 206
|
63.6%
|
# Odds ratio reported only for male participants retained [41] or division of age groups [42].
* Data not divided between those identifying as men and those who identify as women.
a[42] and study protocol [59]. Study launched reported on study website [61].
b Protocol paper describes 3 Stages after recruiting a population-based cohort [59], not described in [42].
c Study contact were text messages the amount, timing and frequency of SMS text messages were decided by participant [46].
d Completers of final postal survey
e Last wave (wave 5) of the study (77 – 87 years old) completers of follow-up, no male specific data given for wave 1-4 [47].
f Protocol paper describes the use of census data at 1996, 2001 and 2006 [62].
g Yearly visits with an option to fill in a complementary questionnaire each month [45].
h Two studies were included WebMOVE and MOVE-SMI [53].
i Sub-study was randomly selected from WebMOVE and MOVE-SMI groups and not stratified by gender [53].
j Due to its longitudinal nature, it is assumed this study has more than 3 visits in the 12-year reported period. No information on number of study visits or contacts was found in Rose et al., [40] or the associated protocol [63].
Table 3: Recruitment strategies and recruitment and retention rates of participants (n.d indicates that no data was available in the published literature)
Recruitment Method
|
Study
|
Recruitment Rate
|
Study Attrition Rate
|
Advertising
|
Crichton et al., 2012 [56]
|
35.7% of total participants (n=30/84) via TV segment
31% of total participants (n=26/84) via Newspaper
|
n=20 males completed the study:
50% completers of 12-month follow up (n=10/20)
(10% Early drop out (n=2/20), Dropouts during study (40% n=8/20))
|
Rose et al., 1976 [40]
|
n.d
|
n= 2,028/2,280 11% attrition. deceased n=54 (2.4%); lost interest n=103 (4.5%); moved away with no further participation n=44 (1.9%); moved away with survey only participation n=51 (2.2%)
|
Méjean et al., 2014 [45]
|
Initial recruitment strategies for NutriNet-Santé Study included television, radio, national and regional newspapers, posters and Internet providing details about the study’s website. A total of 86,652 individuals were recruited (n.d for individuals identifying as male)
|
n.d
|
Letter of Invitation
|
Crichton et al., 2012 [56]
|
84 responders from advertising. Interested potential participants were invited to an information session and pre-study screening. 71 participants screened and deemed eligible (84.5%; n.d for individuals identifying as male)
|
Reported above
|
Cheraghi et al., 2021 [41]
|
All family members were invited for baseline measurements n=15,005 individuals agreed to participate (> 3 years of age, n.d for individuals identifying as male).
|
39.6% of entire cohort after 5th follow up examination
n=10,368 individuals participating in current study. Of these n=4,395 (42.4%) were male (n=1,650 for intervention, n=2,745 for control).
|
Goldberg et al., 2006 [44]
|
GAZEL Study had 20,624 at baseline but 20,328 included in this study. 72.9% of participants identified as male (n=14,827).
|
0.4% attrition rate over 12-year follow up due to leaving the company (n=60), or leaving the study (n=3)
87.2% of participants retained after first annual mail questionnaire returned in 1990 compared to baseline (n.d of numbers of participants or gender)
71.2% retained at the end of the study 12 years later in 2000 (n.d of numbers of participants or gender)
|
Green et al., 2018 [42]
|
Invitation letter signed by GP and information sheet describing aims, nature and how to contact Cam-CAN study was sent to 7,616 eligible individuals. 35.2% (n=2,680) consented. Of those, 43.7% were male (n=1,172).
|
n.d due to single time point reported in this study
|
Irvine et al., 2017 [43]
|
Letter received from GP inviting 47.1% of participants to take part in the study (n=419 contacted out of 889 assessed for eligibility).
|
1.4% (n=1 lost to follow up out of 69 consented participants)
|
Lee et al., 2009 [49]
|
Average recruitment rate of 40%. Recruitment Rate via city and registry: Florence (Primary care; 59.9% n=433/723), Leuven (Electoral; 37.9% n=451/1189), Lodz (City registry; 48.4% n=408/843), Malmö (Population; 44.6% n=409/918), Manchester (Primary care; 37.2% n=396/1064), Santiago (National register; 35.2% n=406/1155), Szeged (Electoral; 24.1% n=431/1789), Tartu (Primary care; 59.2% n=435/735)
|
Average attrition rate of 8.5%. Attrition rate calculated at each site due to death or moved house: Florence (4.4% n=414/433), Leuven (4.4% n=431/451), Lodz (15.7% n=344/408), Malmö (11.3% n=363/409), Manchester (11.1% n=352/396), Santiago (21.2% n=320/406), Szeged (0% n=unknown/431), Tartu (0% n=unknown/435)
|
Oleske et al., 2007 [52]
|
Eligible workers were sent letters from the study principle investigator, plant management and local union official inviting them to the study. N=454 participants joined, of which 79.3% were male (n=360)
|
Overall 31% attrition in 12 months (n=141 drop outs), 83.6% of male participants had complete data (n=168/201), 75.9% of male participants were considered drop out or had missing data collection (n=192/253)
|
Ullman et al.,1998 [55]
|
Mail surveys sent to participants with study information, a University pen and description of incentive once survey was sent back complete. Overall recruitment rate of 76% (n=616/814; n.d for individuals identifying as male). 64.5% responded with no reminder (n=334/518), 18.9% with one reminder (n=98/518), 14.2% with two reminders (n=74/518) and 2.3% with 3 reminders (n=12/518))
32.6% of non-responders (n=32/98) responded after receiving a second letter and $25 cheque.
A substudy of males demonstrated that 93% were retained (201/216). Of these 46.3% of males responded with reminders only (100/216) and a further 7.9% responded with reminders and a cheque (17/216).
|
Overall attrition rate: 10.7% (n=550/616).
|
Markanday et al., 2013 [57]
|
Letter of invitation contained information about the studies purpose, requirements of the participants and time investment required and where study centre was located. 67.1% recruitment rate for those who identified as male (n=1,540/2,296), 77% for those identifying as female (n=1,494/1,938).
|
n.d
|
Méjean et al., 2014 [45]
|
6,556 participated from a total of 15,000 randomly invited participants from the NutriNet-Santé Study. 6,352/6,556 participants available for analysis. Of this sample 1,531 male (24.1%)
|
From those who dropped out of the substudy (n=9,982; defined as participating stopping within 6 months after the inclusion in the cohort), 2,398 identified as male (24%)
|
Recruitment From Hospital or Research Centre
|
Crichton et al., 2012 [56]
|
16.7% of total participants (n=14/84) indicated an interest for future studies at Research Centre. Written advertisement also placed at local hospital n.d on percentages recruited
|
Reported above
|
Janus et al., 1997 [51]
|
n.d
|
7.2% attrition before the first-year assessment (n=15/209) and 28.2% (n=59/209) after the first-year assessment (assumed between years 1-4). 35.4% lost over the 4-year study (n=74/209).
|
Kannisto et al., 2017 [46]
|
33.3% recruitment rate from eligible participants (n=1,139/3,417). Of these, 49.2% (n= 560) who identify as male were randomised.
|
4.8% attrition rate during the intervention period of all participants (n=27/560) with 6 of the 27 withdrawals identifying as male (6/560; 1.1%). During a follow up postal survey 59% attrition rate who identify as male (n=326/589).
|
Invitation Following online intake assessment at clinic or telephone
|
Van Wees et al., 2019 [48]
|
13,658 were eligible and invited. Overall, 12% of eligible individuals were recruited (n=1,705), of those 19% identified as male (n=324)
|
From the total sample, 47.5% attended baseline visit (n=810/1,705), 25.3% attended 3 week visit (n=432/1,705), 24.3% attended 6 month follow-up (n=416/1,705) and 20.2% attended 1 year follow-up (n=344/1,705). Study attrition for the both male and female participants by the 1 year follow up was 79.8%.
|
Recruitment Through Community Venues – Town centre, Workplaces, Community Groups, Football Grounds, Golf Clubs, Library, Shopping Centre
|
Crichton et al., 2012 [56]
|
Written advertisements placed on noticed boards at libraries and shopping centres. n.d on percentages recruited.
|
Reported above
|
Irvine et al., 2017 [43]
|
Field visits through community venues recruited the last 52.9% of participants (n=470 men approached). One participant was recruited for every 11 community venue visits.
|
Reported above
|
Table 4: Barriers to participation and drop-out or refusal rates of participants
Reasons for refusal to participate or drop-out
|
Proportion of participants who refuse participation
|
Reference
|
Appointment non-attendance
|
14.0% of passive refusals
|
[42]
|
2.1% of males that refused to participate
|
[57]
|
1.3% of males were unable to cope with study requirements due to old age
|
[57]
|
Comprehension of the study requirements
|
0.4% of males refused to participate
|
|
Cannot be bothered/not interested
|
27.8% of active refusals
|
[42]
|
13.8% declined to participate
|
[43]
|
66.7% refused to participate
|
[46]
|
39.6% males that refused to participate
|
[57]
|
Time commitment
|
38.9% of active refusals
|
[42]
|
26.3% of males that refused to participate
|
[57]
|
Invasion of privacy
|
0.3% of males that refused to participate
|
[57]
|
Medical
|
15.5% of participants who refused to participate
|
[56]
|
35.6% of participants unable to attend due to illness, 0.2% of participants had limited medical information
|
[42]
|
16.9% of males that refused to participate
|
[57]
|
Unable to contact/no response
|
35.0% of eligible participants
|
[42]
|
17.4% of eligible participants
|
[43]
|
40.7% of invited male participants
|
[49]
|
Psychopathology factors
|
0.8% of males refused to participant in case a medical problem was uncovered
|
[57]
|
Reluctance over medical testing
|
1.1% of males that refused to participate
|
[57]
|
Religious/philosophical reasons
|
0.1% of males that refused to participate
|
[57]
|
Third party involvement
|
62.2% of participants passively refused via a relative, 15.9% of participants passively refused by resident/nursing home
|
[42]
|
18.4% transferred to another ward or discharged from hospital or research nurse forgot to ask
|
[46]
|
Unknown reason/personal reason
|
5.2% of males that refused to participate
|
[57]
|
9.5% of eligible participants in 1968
|
[47]
|
28.6% of active refusals
|
[42]
|
3.2% of eligible participants
|
[43]
|
|
Proportion of participants who were non-completers
|
|
Appointment non-attendance
|
3.2% of non-completers
|
[43]
|
24.7% missed at least one visit by end of study (12 months)
|
[52]
|
Quantitative data- in person visits were difficult to attend due to the distance of the centre
|
[53]
|
Medical
|
12.7% non-completers
|
[56]
|
4.1% of non-completers had a child that had an additional diagnosis
|
[51]
|
Situational (lack of reliable housing, moving, death)
|
1.4% of non-completers
|
[56]
|
95.2% of non-completers
|
[44]
|
1.4% of non-completers from wave 1 (1974) to 47.3% in wave 5 (2011)
|
[47]
|
6.8% to 30.6% of non-completers across 6 different centres
|
[49]
|
25.7% of non-completers moved, 5.4% of families had a child who died
|
[51]
|
21.4% of non-completers died, 17.4% moved away
|
[40]
|
2.3% - 9.4% of non-completers (wave 1-5)
|
[41]
|
Qualitative data- unable to complete exercise or have appropriate meal preparation
|
[53]
|
63.4% of non-completers
|
[54]
|
Inability to adhere to study activities
|
Qualitative data- unable to complete training due to unreliable technology
|
[53]
|
12.7% of non-completers
|
[56]
|
10.1% did not receive allocation of intervention
|
[43]
|
Cannot be bothered/ loss of interest/wanted to withdraw
|
40.8% of non-completers
|
[40]
|
9.5% of non-completers
|
[51]
|
3.2% of non-completers
|
[46]
|
20.6% of non-completers
|
[49]
|
Difficulty to arrange follow-up appointments with participants
|
6.8% of non-completers
|
[51]
|
Missing data/incomplete data
|
6.5% of participants
|
[43]
|
52.5% of participants did not complete the final postal survey, 0.36% of participants did not have available data in the Finnish national Care Register for Health Care
|
[46]
|
15.0% of non-completers
|
[49]
|
3.1% of participants
|
[45]
|
Time commitment
|
5.6% of non-completers
|
[56]
|
17.6% of non-completers
|
[51]
|
Qualitative data- competing demands in personal life, unable to prioritize program participation
|
[53]
|
Lost contact
|
8.5% of non-completers
|
[56]
|
27.8% of non-completers
|
[46]
|
Unknown reason/personal reason
|
7.0% of non-completers
|
[56]
|
4.8% of non-completers
|
[44]
|
4.8% of non-completers
|
[46]
|
21.6% of non-completers
|
[51]
|
20.9% of non-completers from wave 1 (1974) to 24.1% in wave 5 (2011)
|
[47]
|
10.7% of non-completers
|
[55]
|
57.5% of non-completers were lost by 1-year follow up
|
[48]
|
Difficulty in comprehending the study
|
2.7% of non-completers
|
[51]
|
Psychopathology factors
|
Paranoid factor had an elevated but non-significant risk for early drop out (26.9%) Dysphoric Borderline factor put a significant risk for late dropout (15.9%)
|
[50]
|
Qualitative data- side effects from medications for mental health or chronic pain were issues in completing the program. Social anxiety of talking openly to other participants also prohibited some participants interaction.
|
[53]
|
Third party involvement
|
6.8% of non-completers – due to family issues
|
[51]
|
Financial Hardship
|
Qualitative result- participants across all treatment groups found recommendations of what to eat and how to exercise cost prohibitive
|
[53]
|
Technical Issues
|
Qualitative data- difficulties in troubleshooting web-based program after logging in as well as printing physical activity log
|
[53]
|
Table 5: Facilitating factor that improved recruitment and retention of men
Facilitating factor or approach
|
Study details (for either gender)
|
Study details for male specific approaches
|
Advertising through mainstream media, recurrent attention from medical press and general media
|
Crichton et al., 2012 [56], Goldberg et al., 2006 [44], Méjean et al., 2014 [45], Snow et al., 2007 [54]
|
Lee et al., 2009 [49]
|
Annual Cohort Symposium
|
Goldberg et al., 2006 [44]
|
-
|
Delivery of intervention using friendly, relaxed, non-directive style with easily understood information
|
-
|
Irvine et al., 2017 [43]
|
Development of loyal, close relationships in person or one on one check-ins over the phone
|
In person: Janus et al., 1997 [51]
Telephone check-in: Olmos-Ochoa et al., 2019 [53]
|
|
Discussion of reasons for refusal, drop out, or non-attendance
|
Green et al., 2018 [42], Janus et al., 1997 [51], Kannisto et al., 2017 [46], Markanday et al., 2013 [57], Olmos-Ochoa et al., 2019 [53]
|
Lee et al., 2009 [49]
|
Enrolment of spouse or all family members
|
Cheraghi et al., 2021 [41], Janus et al., 1997 [51]
|
Rose et al., 1976 [40]
|
Free medical screening
|
vanWees et al., 2019 [48], Méjean et al., 2014 [45], Crichton et al., 2012 [56]
|
Rose et al., 1976 [40]
|
Incentives/Reimbursement
|
Monetary: Ullman et al.,1998 [55], Crichton et al., 2012 [56], vanWees et al., 2019 [48]
Paid parking: Janus et al., 1997 [51]
Small gifts: Méjean et al., 2014 [45], Oleske et al., 2007 [52], Ullman et al.,1998 [55], Janus et al., 1997 [51]
|
-
|
Invitation to press conferences
|
Méjean et al., 2014 [45]
|
-
|
Maintaining regular contact using one or several methods of contact
|
Crichton et al., 2012 [56], Janus et al., 1997 [51], Kannisto et al., 2017 [46]
|
Irvine et al., 2017 [43]
|
Membership card to study and certificate of completion at each follow-up
|
Méjean et al., 2014 [45]
|
-
|
Participant choice in amount, timing and frequency of participant intervention or location of interview/ Convenience of time/location
|
Timing and frequency: Kannisto et al., 2017 [46],
Time and location of interview: Green et al., 2018 [42]
Convenient timing and/or location: Markanday et al., 2013 [57]
|
Convenient timing and/or location: Irvine et al., 2017 [43], Rose et al., 1976 [40]
|
Participant newsletters
|
Yearly newsletter written from PI: Goldberg et al., 2006 [44]
Annual holiday letter: Janus et al., 1997 [51]
Monthly email with scientific information about health/nutrition: Méjean et al., 2014 [45]
|
Birthday and Christmas cards: Lee et al., 2009 [49]
Study Newsletter: Rose et al., 1976
|
Participant perceptions
|
Aiding research particularly publicly funded research: Méjean et al., 2014 [45]
|
Perceived health benefits: Rose et al., 1976 [40], Irvine et al., 2017 [43]
Satisfaction at being part of ‘health elite’ (Hawthorne Effect): Rose et al., 1976 [40]
|
Payment of wages to employers to attend study visit without loss of income or work penalty
|
Oleske et al., 2007 [52]
|
Rose et al., 1976 [40]
|
Personal notes to participants
|
Ullman et al.,1998 [55]
|
-
|
Prompting participants to update changes in address or phone number by SMS or via post
|
-
|
SMS: Irvine et al., 2017 [43]
Post: Lee et al., 2009 [49]
|
Reminders for questionnaires/ appointments or follow up after non-attendance/non-responders
|
vanWees et al., 2019 [48], Green et al., 2018 [42], Oleske et al., 2007 [52], Ullman et al.,1998 [55], Crichton et al., 2012 [56], Goldberg et al., 2006 [44], Janus et al., 1997 [51], Markanday et al., 2013 [57]
|
Lee et al., 2009 [49]
|
Recruitment via “mass mailing”
|
Snow et al., 2007 [54]
|
-
|
Resources specific to intervention
|
Green et al., 2018 [42], Crichton et al., 2012 [56], Markanday et al., 2013 [57]
|
Irvine et al., 2017 [43], Lee et al., 2009 [49],
|
Screening participants from geographical stable workplaces
|
-
|
Rose et al., 1976 [40]
|
Simple consent process
|
Online : vanWees et al., 2019 [48]
|
Via SMS : Irvine et al., 2017 [43]
|
Study data collected completely online
|
Méjean et al., 2014 [45], Crichton et al., 2012 [56]
|
-
|
Vested personal interest
|
Desire to contribute to chronic disease risk Méjean et al., 2014 [45]
|
Desire to contribute their own weight loss Irvine et al., 2017 [43]
|
Table 6: Quality assessment of paper using the MMAT tool [39].
|
S1
|
S2
|
1.1
|
1.2
|
1.3
|
1.4
|
1.4
|
2.1
|
2.2
|
2.3
|
2.4
|
2.5
|
3.1
|
3.2
|
3.3
|
3.4
|
3.5
|
4.1
|
4.2
|
4.3
|
4.4
|
4.5
|
5.1
|
5.2
|
5.3
|
5.4
|
5.5
|
Study quality
|
Qualitative
|
No qualitative studies were included
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Quantitative RCT
|
Irvine et al., 2017
|
Y
|
Y
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
High
|
Kannisto et al., 2017
|
Y
|
Y
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
High
|
Oleske et al., 2007
|
Y
|
Y
|
|
|
|
|
|
CT
|
CT
|
Y
|
CT
|
CT
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Poor
|
Crichton et al.,2012
|
Y
|
N
|
|
|
|
|
|
CT
|
Y
|
N
|
CT
|
N
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Poor
|
Snow et al., 2007
|
Y
|
Y
|
|
|
|
|
|
CT
|
Y
|
Y
|
CT
|
CT
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Poor
|
Quantitative non-randomised
|
Markanday et al.,2013
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
High
|
Janus et al.,1997
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
|
|
|
|
|
Medium
|
Quantitative descriptive
|
Goldberg et al.,2006
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
Medium
|
Green, et al.,2018
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
Medium
|
Lee et al.,2009
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
High
|
Hamberger et al., 2000
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
High
|
Kelfve et al.,2017
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
High
|
Méjean et al., 2014
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
Medium
|
Rose et al., 1976
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
Medium
|
Ullman et al.,1998
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
CT
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
Poor
|
vanWees et al., 2019
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
|
|
|
|
|
Medium
|
Cheraghi et al., 2021
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
|
|
|
|
High
|
Mixed methods
|
Olmos-Ochoa et al.,2019
|
Y
|
Y
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Y
|
Y
|
Y
|
CT
|
Y
|
Medium
|