Background
Methods
Inclusion and exclusion criteria
Data sources and search methods
# | Search |
---|---|
Existing published strategy for randomised controlled trials | |
1 | (article OR randomized controlled trials).pt. |
2 | Animals/ |
3 | Humans/ |
4 | #2 NOT (2 AND 3) |
5 | #1 NOT #4 |
Cluster design–related terms | |
6 | (cluster$ adj2 randomi$).tw. |
7 | ((communit$ adj2 intervention$) or (communit$ adj2 randomi$)).tw. |
8 | group$ randomi$.tw. |
9 | #6 OR #7 OR #8 |
10 | intervention?.tw. |
11 | Cluster Analysis/ |
12 | Health Promotion/ |
13 | Program Evaluation/ |
14 | Health Education/ |
15 | #10 OR #11 OR #12 OR #13 OR #14 |
16 | #9 OR #15 |
Bayesian search terms | |
17 | bayes$.af. |
18 | #16 AND #17 |
Final search | |
19 | #18 AND #5 |
20 | limit #19 to (randomized controlled trial) |
Reference sifting and quality control
Data extraction
Analysis
Results
R1 | Carabin H, Millogo A, Ngowi HA, et al. Effectiveness of a community-based educational programme in reducing the cumulative incidence and prevalence of human Taenia solium cysticercosis in Burkina Faso in 2011–14 (EFECAB): a cluster-randomised controlled trial. Lancet Glob Heal. 2018;6(4):e411-e425. doi:10.1016/S2214-109X(18)30027-5 |
R2 | Foxcroft DR, Callen H, Davies EL, Okulicz-Kozaryn K. Effectiveness of the strengthening families programme 10-14 in Poland: Cluster randomized controlled trial. Eur J Public Health. 2017;27(3):494-500. doi:10.1093/eurpub/ckw195 |
R3 | Levy BT, Hartz A, Woodworth G, Xu Y, Sinift S. Interventions to Improving Osteoporosis Screening: An Iowa Research Network (IRENE) Study. J Am Board Fam Med. 2009;22(4):360-367. doi:10.3122/jabfm.2009.04.080071 |
R4 | Ngowi HA, Carabin H, Kassuku AA, Mlozi MRS, Mlangwa JED, Willingham AL. A health-education intervention trial to reduce porcine cysticercosis in Mbulu District, Tanzania. Prev Vet Med. 2008;85(1-2):52-67. doi:10.1016/j.prevetmed.2007.12.014 |
R5 | Rahme E, Choquette D, Beaulieu M, et al. Impact of a general practitioner educational intervention on osteoarthritis treatment in an elderly population. Am J Med. 2005;118(11):1262-1270. doi:10.1016/j.amjmed.2005.03.026 |
R6 | Swanson KM, Chen H-T, Graham JC, Wojnar DM, Petras A. Resolution of Depression and Grief during the First Year after Miscarriage: A Randomized Controlled Clinical Trial of Couples-Focused Interventions. J Women’s Heal. 2009;18(8):1245-1257. doi:10.1089/jwh.2008.1202 |
R7 | Van Deurssen E, Meijster T, Oude Hengel KM, et al. Effectiveness of a Multidimensional Randomized Control Intervention to Reduce Quartz Exposure among Construction Workers. Ann Occup Hyg. 2015;59(8):959-971. doi:10.1093/annhyg/mev037 |
R8 | Amza A, Kadri B, Nassirou B, et al. Community risk factors for ocular chlamydia infection in Niger: Pre-treatment results from a cluster-randomized trachoma trial. PLoS Negl Trop Dis. 2012;6(4). doi:10.1371/journal.pntd.0001586 |
R9 | Hovi T, Ollgren J, Savolainen-Kopra C, T. H, J. O. Intensified hand-hygiene campaign including soap-and-water wash may prevent acute infections in office workers, as shown by a recognized-exposure -adjusted analysis of a randomized trial. BMC Infect Dis. 2017;17(1):47. doi:https://doi.org/10.1186/s12879-016-2157-z |
R10 | Barlis P, Regar E, Serruys PW, et al. An optical coherence tomography study of a biodegradable vs. durable polymer-coated limus-eluting stent: A LEADERS trial sub-study. Eur Heart J. 2010;31(2):165-176. doi:10.1093/eurheartj/ehp480 |
R11 | See CW, O’Brien KS, Keenan JD, et al. The effect of mass azithromycin distribution on childhood mortality: Beliefs and estimates of efficacy. Am J Trop Med Hyg. 2015;93(5):1106-1109. doi:10.1111/sjos.12316 |
M1 | Alexander N, Emerson P. Analysis of incidence rates in cluster-randomized trials of interventions against recurrent infections, with an application to trachoma. Stat Med. 2005;24(17):2637-2647. doi:10.1002/sim.2138 |
M2 | Clark AB, Bachmann MO. Bayesian methods of analysis for cluster randomized trials with count outcome data. Stat Med. 2010;29(2):199-209. doi:10.1002/sim.3747 |
M3 | Nixon RM, Duffy SW, Fender GR. Imputation of a true endpoint from a surrogate: Application to a cluster randomized controlled trial with partial information on the true endpoint. BMC Med Res Methodol. 2003;3:1-11. doi:10.1186/1471-2288-3-17 |
M4 | Olsen MK, DeLong ER, Oddone EZ, Bosworth HB. Strategies for analyzing multilevel cluster-randomized studies with binary outcomes collected at varying intervals of time. Stat Med. 2008;27(29):6055-6071. doi:10.1002/sim.3446 |
M5 | Thompson SG, Warn DE, Turner RM. Bayesian methods for analysis of binary outcome data in cluster randomized trials on the absolute risk scale. Stat Med. 2004;23(3):389-410. doi:10.1002/sim.1567 |
M6 | Turner RM, Prevost AT, Thompson SG. Allowing for imprecision of the intracluster correlation coefficient in the design of cluster randomized trials. Stat Med. 2004;23(8):1195-1214. doi:10.1002/sim.1721 |
M7 | Turner RM, Omar RZ, Thompson SG. Modelling multivariate outcomes in hierarchical data, with application to cluster randomised trials. Biometrical J. 2006;48(3):333-345. doi:10.1002/bimj.200310147 |
M8 | Spiegelhalter DJ. Bayesian methods for cluster randomized trials with continuous responses. Stat Med. 2001;20(3):435-452. doi:10.1002/1097-0258(20010215)20:3<435::AID-SIM804>3.0.CO;2-E |
M9 | Kikuchi T, Gittins J. A behavioural Bayes approach for sample size determination in cluster randomized clinical trials. J R Stat Soc Ser C Appl Stat. 2010;59(5):875-888. doi:10.1111/j.1467-9876.2010.00732.x |
M10 | Turner RM, Thompson SG, Spiegelhalter DJ. Prior distributions for the intracluster correlation coefficient, based on multiple previous estimates, and their application in cluster randomized trials. Clin Trials. 2005;2(2):108-118. doi:10.1191/1740774505cn072oa |
M11 | Turner RM, Omar RZ, Thompson SG. Constructing intervals for the intracluster correlation coefficient using Bayesian modelling, and application in cluster randomized trials. Stat Med. 2006;25(9):1443-1456. doi:10.1002/sim.2304 |
M12 | Uhlmann L, Jensen K, Kieser M. Bayesian network meta-analysis for cluster randomized trials with binary outcomes. Res Synth Methods. 2016;8(October 2015):236-250. doi:10.1002/jrsm.1210 |
M13 | Turner RM, Omar RZ, Thompson SG. Bayesian methods of analysis for cluster randomized trials with binary outcome data. Stat Med. 2001;20(3):453-472. doi:10.1002/1097-0258(20010215)20:3<453::AID-SIM803>3.0.CO;2-L |
C1 | Peters TJ, Richards SH, Bankhead CR, Ades AE, Sterne JAC. Comparison of methods for analysing cluster randomized trials: An example involving a factorial design. Int J Epidemiol. 2003;32(5):840-846. doi:10.1093/ije/dyg228 |
C2 | Pacheco GD, Hattendorf J, Colford JM, Mäusezahl D, Smith T. Performance of analytical methods for overdispersed counts in cluster randomized trials: Sample size, degree of clustering and imbalance. Stat Med. 2009;28(24):2989-3011. doi:10.1002/sim.3681 |
C3 | Ma J, Thabane L, Kaczorowski J, et al. Comparison of Bayesian and classical methods in the analysis of cluster randomized controlled trials with a binary outcome: The community hypertension assessment trial (CHAT). BMC Med Res Methodol. 2009;9(1). doi:10.1186/1471-2288-9-37 |
Demographics
N (%) unless otherwise stated | Total (N = 11) | Primary (N = 7) | Secondary (N = 4) |
---|---|---|---|
Year of publication | |||
Pre 2005 | 0 (0.0) | 0 (0.0) | 0 (0.0) |
2005–2012 | 6 (54.5) | 4 (57.1) | 2 (50.0) |
Post 2012 | 5 (45.5) | 3 (42.9) | 2 (50.0) |
Location of first authora | |||
UK | 2 (18.2) | 1 (14.3) | 1 (25.0) |
US/Canada | 5 (45.5) | 4 (57.1) | 1 (25.0) |
Europe excl. UK | 3 (27.3) | 1 (14.3) | 2 (50.0) |
Australia/New Zealand | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Africa | 2 (18.2) | 1 (14.3) | 1 (25.0) |
Asia | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Other | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Location of studya | |||
UK | 1 (9.1) | 0 (0.0) | 1 (25.0) |
US/Canada | 3 (27.3) | 3 (42.9) | 0 (0.0) |
Europe excl. UK | 4 (36.4) | 2 (28.6) | 2 (50.0) |
Australia/New Zealand | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Africa | 4 (36.4) | 2 (28.6) | 2 (50.0) |
Asia | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Other | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Target sample size; mean (SD) [range] | N/A | N = 3b 1466.7 (1868.6) [120, 3600] | N/A |
Target number of clusters; mean (SD) [range] | N/A | N = 2c 200.0 (198.0) [60, 340] | N/A |
Recruited Sample Size; mean (SD) [range] | N = 11 10898.5 (19816.1) [116, 66204] | N = 7 2484.6 (3700.1) [116, 9928] | N = 4 25662.8 (28762.5) [683, 66204] |
Recruited Number of Clusters; mean (SD) [range] | N = 11 58.8 (95.6) [5, 341] | N = 7 69.1 (121.6) [5, 341] | N = 4 40.8 (13.2) [21, 48] |
Randomisation unit | |||
Medical facility | 1 (9.1) | 1 (14.3) | 0 (0.0) |
Village/community/district | 6 (54.5) | 4 (57.1) | 2 (50.0) |
Organisation | 1 (9.1) | 1 (14.3) | 0 (0.0) |
Couple | 1 (9.1) | 1 (14.3) | 0 (0.0) |
Individual | 1 (9.1) | 0 (0.0) | 1 (25.0) |
Working unit (office) | 1 (9.1) | 0 (0.0) | 1 (25.0) |
Primary outcome type | |||
Binary | 9 (81.8) | 5 (71.4) | 4 (100.0) |
Continuous | 2 (18.2) | 2 (28.6) | 0 (0.0) |
Statistician involvement | 8 (72.7) | 5 (71.4) | 3 (75.0) |
Statistician association | |||
Clinical trials unit | 1 (12.5) | 0 (0.0) | 1 (33.3) |
Academic statistical department | 7 (87.5) | 5 (100.0) | 2 (66.6) |
Commercial pharmaceutical company | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Clinical research organisation | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Other | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Journal endorsement of CONSORT guidelines | |||
High | N/A | 3 (42.9) | N/A |
Medium | N/A | 1 (14.3) | N/A |
Low | N/A | 0 (0.0) | N/A |
None | N/A | 3 (42.9) | N/A |
Reporting quality
Reporting quality criteria N (%) | Total (N = 7) | Year of publication | Journal endorsement of CONSORT guidelines | Statistician involvement | |||
---|---|---|---|---|---|---|---|
2012 or earlier (N = 4) | 2013 onwards (N = 3) | High/medium (N = 4) | Low/none (N = 3) | Yes (N = 5) | No (N = 2) | ||
Description of sample size method | 4 (57.1) | 2 (50.0) | 2 (66.7) | 2 (50.0) | 2 (66.7) | 2 (40.0) | 2 (100.0) |
Was clustering clearly accounted for in sample size calculation | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Specification of the required number of clusters | 2 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) |
Specification of the assumed cluster size | 2 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) | 1 (50.0) |
Specification of whether equal or unequal cluster sizes are assumed | 1 (25.0) | 1 (50.0) | 0 (0.0) | 0 (0.0) | 1 (50.0) | 0 (0.0) | 1 (50.0) |
Variability in cluster size accounted for | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Specification of the ICC used for the sample size | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Indication of the uncertainty of the ICC | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
Accounted for the uncertainty in the ICC | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
Other CONSORT metrics | |||||||
Details of how clustering was accounted for in the analysis | 6 (85.7) | 4 (100.0) | 2 (66.7) | 4 (100.0) | 2 (66.7) | 5 (100.0) | 1 (50.0) |
Specification of the number of clusters randomised | 7 (100.0) | 4 (100.0) | 3 (100.0) | 4 (100.0) | 3 (100.0) | 5 (100.0) | 2 (100.0) |
Specification of the number of clusters receiving intended treatment | |||||||
Explicit | 5 (71.4) | 3 (75.0) | 2 (66.7) | 4 (100.0) | 1 (33.3) | 4 (80.0) | 1 (50.0) |
Implied | 2 (28.6) | 1 (25.0) | 1 (33.3) | 0 (0.0) | 2 (66.7) | 1 (20.0) | 1 (50.0) |
Specification of the number of clusters analysed for the primary outcome at the primary endpoint | |||||||
Explicit | 2 (28.6) | 1 (25.0) | 1 (33.3) | 2 (50.0) | 0 (0.0) | 2 (40.0) | 0 (0.0) |
Implied | 5 (71.4) | 3 (75.0) | 2 (66.7) | 2 (50.0) | 3 (100.0) | 3 (60.0) | 2 (100.0) |
Details of cluster-level losses and exclusions | |||||||
Explicit | 3 (42.9) | 2 (50.0) | 1 (33.3) | 2 (50.0) | 1 (33.3) | 2 (40.0) | 1 (50.0) |
Implied | 4 (57.1) | 2 (50.0) | 2 (66.7) | 2 (50.0) | 2 (66.7) | 3 (60.0) | 1 (50.0) |
Details of individual-level losses and exclusions | 4 (57.1) | 2 (50.0) | 2 (66.7) | 2 (50.0) | 2 (66.7) | 2 (40.0) | 2 (100.0) |
Individual-level baseline characteristics presented | 7 (100.0) | 4 (100.0) | 3 (100.0) | 4 (100.0) | 3 (100.0) | 5 (100.0) | 2 (100.0) |
Cluster-level baseline characteristics presented | 2 (28.6) | 2 (50.0) | 0 (0.0) | 1 (25.0) | 1 (33.3) | 1 (20.0) | 1 (50.0) |
Coefficients of intracluster correlation provided for primary outcomes | |||||||
All | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Some | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Coefficients of intracluster correlation provided for secondary outcomes | |||||||
All | 0 (0.0)a | 0 (0.0) | 0 (0.0)a | 0 (0.0) | 0 (0.0)a | 0 (0.0) | 0 (0.0)a |
Some | 0 (0.0)a | 0 (0.0) | 0 (0.0)a | 0 (0.0) | 0 (0.0)a | 0 (0.0) | 0 (0.0)a |
P-values provided for baseline comparisons | 5 (71.4) | 3 (75.0) | 2 (66.7) | 3 (75.0) | 2 (66.7) | 3 (60.0) | 2 (100.0) |
Clustering accounted for in the calculation of the p-values | |||||||
Yes | 1 (20.0) | 1 (33.3) | 0 (0.0) | 1 (33.3) | 0 (0.0) | 1 (33.3) | 0 (0.0) |
Unclear | 1 (20.0) | 1 (33.3) | 0 (0.0) | 1 (33.3) | 0 (0.0) | 1 (33.3) | 0 (0.0) |
Use of Bayesian methodology
N (%) | Total (N = 11) | Primary (N = 7) | Secondary (N = 4) |
---|---|---|---|
Sample Size (used) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Sample Size (discussed) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
Analysis (used) | 11 (100.0) | 7 (100.0) | 4 (100.0) |
Priors used | |||
Informative | 2 (18.2)a | 1 (14.3)a | 1 (25.0) |
Weakly Informative | 1 (9.1) | 1 (14.3) | 0 (0.0) |
Non-informative | 5 (45.5)a | 3 (42.9)a | 2 (50.0) |
Unspecified | 4 (36.4) | 3 (42.9) | 1 (25.0) |
Analysis (discussed) | N/A | N/A | N/A |