Table 4 Summary of results from included studies
Decision rule | Study design | Study | Study setting | Sample size | Intervention vs. control | Results |
|---|---|---|---|---|---|---|
Use of rules - assessed by documentation of clinical features consistent with rules | ||||||
OAR | RCT | Tajmir (2017)a | Urgent care affiliated with quaternary care, academic hospital, Boston | 613 patients (258 in pre-intervention period and 374 in post-intervention period) | Reminders vs. no intervention | Control vs. Intervention Ankle Rules 231/374 (62%) vs. 239/258 (93%), p = 0.02 Foot rules 238/374 (64%) vs. 209/258 (81%), p < 0.01 |
Auleley (1997)b | Emergency departments of 5 Paris University teaching hospitals | 4129 patients (1992 in the intervention and 2137 in the control group) | Educational meetings + educational materials vs. no intervention | Intervention phase (no data for control ) Intervention group (no n reported) 93% | ||
Single-arm trial | Silveira (2016) | Emergency Department, 793-bed, quaternary care, academic hospital, Boston | 460 ED visits for 45 patients (205 in pre-intervention period and 255 in post-intervention period) | Reminders vs. no intervention | Pre vs. Post-intervention 229/410 (56%) vs. 488/510 (96%), p < 0.001 | |
Ashurst (2014) | One tertiary care centre & one urban Emergency department in Bethlehem, Pennsylvania | 60 patients (30 in pre-intervention period and 30 in post-intervention period) | Educational meetings + educational materials vs. no intervention | At triage (pre- vs. post-intervention) 1/30 (3%) vs. 2/30 (7%), p < 0.001 After triage (pre- vs. post-intervention) 6/30 (20%) vs. 25/30 (83%), p < 0.001 | ||
Stiell (1995) | Emergency departments across several regions | 1,276,288 patients (30 in pre-intervention period and 6489 in post-intervention period) | Educational meetings + educational materials + patient-mediated interventions vs. no intervention | Intervention phase (no pre-intervention data) Post-intervention - Ankle n = 4768/5003 (95%) - Foot n = 4753/5003 (95%) | ||
Bessen (2009) | Emergency departments of a tertiary teaching hospital and a community hospital, Adelaide | 1561 patients (459 in pre-intervention period and 1102 in post-intervention period) | Educational meetings + local consensus processes + local opinion leaders + reminders + audit and feedback vs. no intervention | Request forms (pre- vs. post-intervention) - Tertiary hospital 84/205 (41%) vs. 588/707 (83%), p < 0.001 - Community hospital 76/223 (34%) vs. 147/225 (65%), p < 0.001 Case notes (pre- vs. post-intervention) - Tertiary hospital 123/214 (58%) vs. 767/810 (95%), p < 0.001 - Community hospital 216/244 (52%) vs. 231/286 (81%), p < 0.001 | ||
OKR | Single-arm trial | Atkinson (2004) | Emergency department of a district general hospital, Luton | 130 patients (58 in pre-intervention period and 72 in post-intervention period) | Reminders vs. no intervention | Pre vs. Post-intervention 44/58 (76%) vs. 67/72 (93%) (p-value not provided) |
O’Sullivan (2006) | Emergency Department in Cork | 79 patients (43 in pre-intervention period and 36 in post-intervention period) | Educational meetings + reminders vs. no intervention | Pre vs. Post-intervention 10/29 (35.5%) vs. 14/23 (61%), p = 0.05 | ||
NEXUS | Single-arm trial | Rostas (2015) | Combined adult and pediatric Level I trauma centre, Richmond, Virginia | 233 patients (128 in pre-intervention period and 105 in post-intervention period) | Clinical practice guideline vs. no intervention | Pre vs. Post-intervention 46/54 (85%) vs. 24/30 (80%) (p-value not provided) |
CCHR | RCT | Stiell (2010)b | 12 hospital emergency department sites in the Canadian provinces of Ontario, Alberta, and British Columbia (6 teaching hospitals and 6 community hospitals) | 4531 patients (2580 in intervention group and 1951 in control group) | Clinical practice guidelines + educational materials + educational meetings + reminders vs. no intervention | Post-intervention (no data for control) Intervention group 909/1166 (78%) |
Single-arm trial | Helms (2023) | Emergency department of a 282-bed acute care hospital, located in a suburban community | 600 medical records | Interactive educational meetings + educational materials + Clinical Practice Guidelines + reminders | Pre vs. Post-intervention 171/264 (64.6%) vs. 249/336 (74.3%), p = 0.01 | |
Indicated and non-indicated imaging | ||||||
OAR | RCT | Auleley (1997) | Emergency departments of 5 Paris University teaching hospitals | 4129 patients (1992 in the intervention and 2137 in the control group) | Educational meetings + educational materials vs. no intervention | Intervention groupc Indicated radiographs (n not reported) 98.5% Non-indicated radiographs (n not reported) 20.5% |
Single-arm trial | Stiell (1995) | Emergency departments across several regions | 1,276,288 patients (30 in pre-intervention period and 6489 in post-intervention period) | Educational meetings + educational materials + patient-mediated interventions vs. no intervention | Non-indicated radiographs (post-intervention data) (no pre-intervention data) (n not reported) 4.9% | |
OKR | Single-arm trial | Atkinson (2004) | Emergency department of a district general hospital, Luton | 130 patients (58 in pre-intervention period and 72 in post-intervention period) | Reminders vs. no intervention | Indicated radiographs (pre vs. post) 24/28 (86%) vs. 36/37 (97%) (p-value not provided) Non-indicated radiographs (pre vs. post) 10/30 (33%) vs. 4/35 (11%), p = 0.016 |
O’Sullivan (2006) | Emergency Department in Cork | 79 patients (43 in pre-intervention period and 36 in post-intervention period) | Educational meetings + reminders vs. no intervention | Indicated radiographs (pre vs. post) 10/11 (91%) vs. 14/15 (93%) (p-value not provided) Non-indicated radiographs (pre vs. post) 19/32 (59%) vs. 9/21 (43%), p = 0.05 | ||
NEXUS | Single-arm trial | Rostas (2015) | Combined adult and pediatric Level I trauma centre, Richmond, Virginia | 233 patients (128 in pre-intervention period and 105 in post-intervention period) | Clinical practice guideline vs. no intervention | Indicated CT (pre vs. post) 46/76 (61%) vs. 24/64 (38%), p = 0.01) Non-indicated CT (pre vs. post) 9/51 (18%) vs. 6/41 (15%) (p-value not provided) |
Baker (2020) | Emergency department | 445 patients (42 in pre-intervention period and 403 in post-intervention period) | Educational meetings and materials + reminders + audit and feedback vs. no intervention | Indicated radiographs (pre vs. post) 23/42 (55%) vs. 313/403 (78%) (p-value not provided) Non-indicated radiographs (pre vs. post) 19/42 (45%) vs. 90/403 (22%), p = 0.002 | ||
Griffith (2014)d | Level 1 adult trauma centre, Detroit, Michigan | 895 patients (507 in pre-intervention period and 388 in post-intervention period) | Interactive educational Meeting + audit and feedback and continuous quality improvement + reminders vs. no intervention | Indicated radiographs (pre vs. post) 426/507 (84%) vs. 339/388 (87%) (p-value not provided) Non-indicated radiographs (pre vs. post) 81/507 (16%) vs. 49/376 (13%), p = 0.2 | ||
CCR | Single-arm trial | Griffith (2014)d | Level 1 adult trauma centre, Detroit, Michigan | 895 patients (507 in pre-intervention period and 388 in post-intervention period) | Interactive educational meeting + audit and feedback and continuous quality improvement + reminders vs. no intervention | Indicated radiographs (pre vs. post) 297/416 (71%) vs. 249/320 (78%) (p-value not provided) Non-indicated radiographs (pre vs. post) 119/416 (29%) vs. 71/312 (23%) (p-value not provided) |
Overall imaging use | ||||||
OAR | RCT | Tajmir (2017)a | Urgent care affiliated with quaternary care, academic hospital, Boston | 613 patients (258 in pre-intervention period and 374 in post-intervention period) | Reminders vs. no intervention | Radiography use (control vs. intervention) Ankle 183/374 (49%) vs. 166/258 (64%), p < 0.01 Foot 202/374 (54%) vs. 141/258 (55%), p = 0.95 Both ankle and foot 59/374 (16%) vs. 65/258 (25%), p = 0.0039 |
Auleley (1997) | Emergency departments of 5 Paris University teaching hospitals | 4129 patients (1992 in the intervention and 2137 in the control group) | Educational meetings + educational materials vs. no intervention | Radiography use (control vs. intervention) Pre-intervention 1115/1132 (99%) vs. 1064/1086 (98%) (p-value not provided) Intervention 996/1005 (99%) vs. 691/906 (76%) (p-value not provided) Post-interventionc No control data reported vs. 707/851 (83.1%) (p-value not provided) | ||
Quasi experimental controlled trial | Holroyd (2004) | Emergency departments of 4 major hospitals in Edmonton | 6398 patients (3041 in the intervention and 3367 in the control group) | Int 1: Educational materials + interactive educational meetings Int 2: Educational meetings + Clinical Practice Guidelines Int 3: Educational meetings + audit and feedback + monitoring the performance of the delivery of healthcare VS. Int 1: Educational materials + interactive educational meetings | Radiography use (control vs. intervention) Baseline period (p-value not provided) - Total x-rays taken 571/623 (92%) vs. 444/481 (92%) - Dual x-ray, both ankle and foot 91/571 (16%) vs. 97/444 (22%) - Single x-ray, ankle, or foot 480/571 (84%) vs. 347/444 (78%) End of Intervention 1 (p-value not provided) - Total x-rays taken 884/961 (92%) vs. 786/847 (93%) - Dual x-ray, both ankle and foot 133/884 (15%) vs. 131/786 (17%) - Single x-ray, ankle, or foot 751/884 (85%) vs. 655/786 (83%) End of Intervention 2 (p-value not provided) - Total x-rays taken 697/760 (92%) vs. 676/719 (94%) - Dual x-ray, both ankle and foot 129/697 (19%) vs. 112/676 (17%) - Single x-ray, ankle, or foot 568/697 (75%) vs. 564/676 (78%) End of Intervention 3 - Total x-rays taken 925/1023 (90%) vs. 919/984 (93%) - Dual x-ray, both ankle and foot 162/925 (18%) vs. 169/919 (18%) - Single x-ray, ankle, or foot 763/925 (75%) vs. 750/919 (76%) | |
Single-arm trial | Silveira (2016) | Emergency Department, 793-bed, quaternary care, academic hospital, Boston | 460 ED visits for 457 patients (205 in pre-intervention period and 255 in post-intervention period) | Reminders vs. no intervention | Radiography use (pre vs. post) Ankle 160/205 (78%) vs. 197/255 (77%), p = 0.839 Foot 103/205 (50%) vs. 117/255 (46%), p = 0.352 Ankle or foot 198/205 (97%) vs. 242/255 (95%), p = 0.379 | |
Ashurst (2014) | One tertiary care centre & one urban Emergency department in Bethlehem, Pennsylvania | 60 patients (30 in pre-intervention period and 30 in post-intervention period) | Educational meetings + educational materials vs. no intervention | Radiography use (pre- vs. post-intervention) 27/30 (90%) vs. 24/30 (80%), p = 0.472 | ||
Bessen (2009) | Emergency departments of a tertiary teaching hospital and a community hospital, Adelaide | 1561 patients (459 in pre-intervention period and 1102 in post-intervention period) | Educational meetings + local consensus processes + reminders + audit and feedback vs. no intervention | Radiography use (pre- vs. post-intervention) Tertiary hospital 206/215 (96%) vs. 709/813 (87%), p < 0.001 Community hospital 223/244 (91%) vs. 228/289 (79%), p < 0.001 | ||
Cameron (1999) | 63 Ontario hospitals, including 5 primarily smaller community hospitals (Group A), 5 primarily larger community hospitals (Group B), 4 larger community hospitals and 1 teaching hospital (Group C) | 407 clinicians seeing 1648 patients (830 in pre-intervention period and 818 in post-intervention period) | Educational meetings + educational materials vs. no intervention | Radiography use (pre- vs. post-intervention) Group A hospitals Ankle 119/162 (73%) vs. 141/190 (75%), p = 0.81 Foot 25/162 (15%) vs. 38/190 (20%), p = 0.27 Group B hospitals Ankle 176/241 (73%) vs. 190/235 (81%), p = 0.050 Foot n = 45/241 (19%) vs. 52/235 (22%), p = 0.20 Group C hospitals Ankle 181/240 (75% ) vs. 136/208 (65%), p = 0.022 Foot 46/240 (19%) vs. 44/208 (21%), p = 0.64 | ||
Gwilym (2003) | Emergency department, Northampton General Hospital, Northampton | 207 patients (106 in pre-intervention period and 101 in post-intervention period) | Educational meetings + educational materials vs. no intervention | Radiography use (pre- vs. post-intervention) 75/106 (71%) vs. 57/101 (56%), p < 0.05 | ||
Stiell (1995) | Emergency departments across several regions | 1,276,288 patients (30 in pre-intervention period and 6489 in post-intervention period) | Educational meetings + educational materials + patient-mediated interventions vs. no intervention | Radiography use (pre- vs. post-intervention) 5207/6288 (83%) vs. 3955/6489 (61%), p < 0.001 | ||
OKR | Single-arm trial | Atkinson (2004) | Emergency department of a district general hospital, Luton | 130 patients (58 in pre-intervention period and 72 in post-intervention period) | Reminders vs. no intervention | Radiography use (pre vs. post) 34/58 (59%) vs. 40/72 (56%), p = 0.726 |
O’Sullivan (2006) | Emergency Department in Cork | 79 patients (43 in pre-intervention period and 36 in post-intervention period) | Educational meetings + reminders vs. no intervention | Radiography use (pre- vs. post-intervention) 29/43 (67%) vs. 23/36 (64%), p = 0.05 | ||
Wigder (1999) | Emergency Department, Lutheran General Hospital, Chicago, Illinois | 27 physicians seeing 362 patients (171 in pre-intervention period and 191 in post-intervention period) | Educational meetings + audit and feedback + educational materials + reminders vs. no intervention | Radiography use (pre- vs. post-intervention) 157/171 (92%) vs. 135/191 (71%) (p-value not provided) | ||
NEXUS | Single-arm trial | Rostas (2015) | Combined adult and pediatric Level I trauma centre, Richmond, Virginia | 233 patients (128 in pre-intervention period and 105 in post-intervention period) | Clinical practice guideline vs. no intervention | CT use (pre vs. post) 55/128 (43%) vs. 30/105 (29%), p = 0.01 |
CCR | Single-arm trial | Kerr (2005) | Emergency Department of a community, teaching hospital in Melbourne | 211 patients with head and neck injury (98 in pre-intervention period and 113 in post-intervention period) | Educational meetings + reminders + Clinical practice guideline vs. no intervention | Radiography use (pre- vs. post-intervention) 66/98 (67.3%) vs. 57/113 (50.4%), p = 0.0187 |
CCHR | RCT | Stiell (2010) | 12 hospital emergency department sites in the Canadian provinces of Ontario, Alberta, and British Columbia (6 teaching hospitals and 6 community hospitals) | 4531 patients (2580 in intervention group and 1951 in control group) | Clinical practice guidelines + educational materials + educational meetings + reminders vs. no intervention | CT use (control vs. intervention) Pre-intervention 591/876 (68%) vs. 659/1049 (63%) (p-value not provided) Post-intervention 797/1075 (74%) vs. 1167/1531 (76%) (p-value not provided) |
Andruchow (2021) | Five urban acute care sites Calgary, Alberta: - one level one trauma centre, - three community EDs and - one urgent care centre | 5687 patients (3085 in intervention group and 2602 in control group) | Reminders + audit and feedback or continuous quality improvement vs. no intervention | CT use (pre- vs. post-intervention) Intervention 2133/5136 (41.5%) vs. 1227/3085 (39.8%), p = 0.31 Pre-intervention 1979/4614 (42.9%) vs. 1112/2602 (42.7%) p-value not provided) | ||
Single-arm trial | Kim (2021) | Emergency Department of a large Midwestern academic medical centre in California | 697 adult patients (467 in pre-intervention period and 230 in post-intervention period) | Educational meetings + education materials + reminders vs. no intervention | Radiography use (pre- vs. post-intervention) 399/467 (85.4%) vs. 169/230 (73.4%) (p-value not provided) | |
Mean knowledge scores of the rules | ||||||
OAR | RCT | Gravel (2010) | A paediatric ED located in an urban, tertiary care, Level 1 trauma centre, Montreal, Quebec | 190 clinicians (95 in the control group and 95 in the intervention group) Pre-intervention, N = 190 Control group N = 95 Intervention group N = 95 At 3 weeks, N = 181 Control group N = 92 Intervention group N = 89 At 5 to 9 months, n = 138 Control group N = 68 Intervention group N = 70 | Clinical practice guideline vs. standard description of OAR. | Mean questionnaire scores for knowledge of rules, 0–13 for Ankle Rules and 0–10 for Foot Rules (a higher score means better knowledge) (Control vs. intervention) Preintervention - Ankle rules 3.8 (95% CI: 3.0 to 4.6) vs. 3.5 (95% CI: 2.8 to 4.2) - Foot rules 2.4 (95% CI: 1.8 to 3.0) vs. 2.3 (95% CI: 1.7 to 2.9) At 3 weeks - Ankle rules 10.2 (95% CI: 9.6 to 10.9) vs. 10.9 (95% CI: 10.3 to 11.6) - Foot rules 7.5 (95% CI: 6.9 to 8.0) vs. 7.6 (95% CI: 7.0 to 8.1) At 5 to 9 months - Ankle rules 8.9 (95% CI: 8.3 to 9.5) vs. 10.1 (95% CI: 9.5 to 10.6) - Foot rules 6.5 (95% CI: 5.9 to 7.1) vs. 7.8 (95% CI: 7.2 to 7.3) |
CCHR | Single-arm trial | Zakhari (2016) | Acute Care Academic Facility, New York | 100 clinicians | Interactive educational meetings + reminders vs. no intervention | Mean knowledge scores of CCHR, 0-100% (pre- vs. post-intervention) 49% vs. 89% (p-value not provided) |