Table 1 Characteristics and Outcomes of Included Studies
From: Strategies for improving physician documentation in the emergency department: a systematic review
Author Date Country | Study designa | Setting Participants | Intervention(s) Duration | Intervention & Control Group Details | Outcomes of Interest | Results | Downs & Black Quality Score |
|---|---|---|---|---|---|---|---|
Carter et al. USA (2009) | PPN | • Teaching hospital • Residents (R2, R3) | Multiple Intervention (audit/feedback, education, and reminders) 12 weeks | Intervention group n = 24 1-h lecture to 18/24 residents Pocket card and lecture handouts to 24/24 residents and biweekly newsletters. Physicians received weekly case specific chart audit/feedback. Control group (n = 24) Usual electronic documentation program | • Chart level, based on complexity of decision making and detail of history and physical. • RVU (relative value units). • Billings/hr. | • Intervention resulted in more complex charting (27% vs 19%, p < .01) and fewer mid-level charts (p < .01). • RVUs increased with intervention (3.71 vs 3.17, p < .01). • Billings increased with intervention ($354.08 vs $303.79, p < .01). | 19/27 |
Cole & Counselman USA (1995) | RFUP | • Teaching hospital • Residents & Physicians | Dictation 50 weeks | Intervention group n = 94 Dictated report Dictation services available for 8 h per day, alternating between day and evening shifts. Control group n = 108 Usual paper charts | Mean number of 28 critical items present in report. | Significant mean increase in the number of critical items reported (19.6 vs 15.8; p < .01). | 19/27 |
De Winter et al. Belgium (2011) | PPN | • Teaching hospital • General internist and internal medicine trainees | Reminder 14 weeks | Intervention group n = 924 A limited questions list to encourage collection of data on patients’ prescription and non-prescription medications. Pharmacists interviewed patients to collect complete medication histories for gold standard comparison. Control group n = 798 Usual paper chart | Proportion of drugs omissions in physician history compared to pharmacy-technician gold standard history taking. | Significant decrease in the proportion of drug omissions (9% vs. 17%, p < .001). | 20/27 |
Dexter et al. UK (2008) | PFUP | • Otolaryngology Emergency Clinic • “Doctors” | Multiple Intervention (education and template) Not specified | Intervention group n = 140 Proformas to encourage documentation of patient information. Advice provided on how to improve handwriting. Case notes audited using an ANKLe (Adjusted Note Keeping and Legibility) scoring system. Control group n = 140 Usual documentation | Legibility, content, and ANKLe (Adjusted Note Keeping and Legibility) scores. | Significant improvements in mean ANKLe scores for note content (17.2 vs. 16.0, p < 0.05), legibility (3.02 vs. 2.96, p < 0.05) and overall ANKLe score (20.24 vs. 18.95, p < 0.05). | 15/27 |
Goodyear et al. UK (1995) | CSC | • Emergency Department • “Junior doctors” | Template 30 weeks | Intervention group n = 100 Pre-printed pediatric admission assessment forms. Control group n = 100 Usual handwritten medical records. | Mean numbers of 25 core clinical details recorded: mean number of words per clerking. | Significant increase in number of core clinical details recorded - 24 recorded with intervention vs. 17.6 (p < 0.001) and words per clerking 144 for intervention vs. 184 (p < 0.001). | 12/27 |
Hanson et al. UK (1994) | TSS | • 2 Teaching hospitals • House Officers | Audit/Feedback 19 weeks | Intervention group n = 420(Feedback 1); 429(Feedback 2); 244 (Final audit – weeks 20–24) Phase I: Feedback at 6 weeks in form of individual audit/feedback and group discussion. Phase II: Feedback at week 11. Further audit during weeks 11–16. Post-intervention final audit weeks 20–24. Control group n = 401 No feedback. Baseline audits. Usual paper charts. | • Proportion of head injury charts documenting GCS (Glascow Coma Scale). • Proportion of charts documenting diagnostic coding for all patients. | • Significant improvement in GCS documentation for both hospitals for patients with head injuries during all phases of the study – (80% Feedback 1, 88% Feedback 2, 90% Final Audit vs 40% at baseline). • Significant improvement in diagnostic coding for Hospital A from baseline (p < .008). | 18/27 |
Heidt & Griffey USA (2012) | PPN | • Teaching hospital • Emergency Physicians | Audit/Feedback 12 weeks | Intervention group n = 382 Individualized email feedback from coders to physicians whose charts lacked sufficient documentation to warrant the inclusion of critical care billing codes Control group n = 501 No feedback. | Proportion of ICU (intensive care unit) admissions that documented critical care time. | Significant increase in the number of charts documenting critical care time (64% vs 18%, p < .001). | 10/27 |
Humphreys et al. USA (1992) | CSC | • Teaching hospital • Internal medicine housestaff and ED physicians | Template 31 weeks | Intervention group n = 99 Preformatted chart for obstetric or gynaecological problems. Control group n = 60 Standard blank charts. | Proportion of ICU (intensive care unit) admissions that documented critical care time in the emergency room. | Significant increase in documentation of critical care time (243/382 (64%) vs 88/501 (18%) - p < 0.001). | 22/27 |
Kondziolka et al. Canada (1989) | PPN | • Regional trauma unit • Physicians | Template Not specified | Intervention group n = 100 Neurotrauma Assessment Record templates with 32 information parameters. Control group n = 100 Usual paper chart | Proportion of charts with each of 32 assessed items. | Significant improvement in the recording of elements including incident time and transfer, and medical history (p < .001), and a significant decrease in recording of treatment plans (p < .001). | 19/27 |
Marill et al. USA (1999) | RCT | • ER trauma centre • Physicians, Residents & Medical Students | Template Approx. 2.5 weeks (16 days) | Intervention group n = 657 Commercial templates-guided medical documentation system for all patients presenting to ERs during a 16 day period. Control group n = 570 Usual paper chart | • Emergency physician total treatment and evaluation time. • Total professional bill and physician satisfaction with documentation method. | • Non-significant reduction of 4.6 min in treatment time (95% confidence interval [CI], −9.2 to 18.3). • Significant mean increase in total billing ($137.40 vs $107.80; 95% CI for difference - $22.20 to $37.00). | 25/27 |
O’Connor et al. New Zealand (2001) | PPN | • Non-teaching rural hospital • Physicians | Template 2 weeks | Intervention group n = 96 Preformatted emergency department charts with 8 key content items Control group n = 137 Usual paper charts | • Median number of parameters filled in for each chart, out of 8. • Proportion of charts recording each of 8 parameters. | • Significant mean increase in the number of parameters documented in each chart (8 vs 7, p = .005). • Significant positive change in the recording of one parameter – Physician Name (52% vs 18%, p < .0001). | 19/27 |
Otillo et al. USA (2014) | PPC | • Academic children’s hospital • Pediatric residents | Education 112 weeks | Intervention group n = 157 One-hour lecture Control group n = 145 No education | Proportion of charts with documentation of 3 specific findings. | No change in right lower quadrant tenderness documentation (for example): 43.9% vs. 35.9%, 95% CI -19 to + 3 | 20/27 |
Schnieden & Good Australia (1996) | PPN | • Emergency department • House Officers & Physicians | Template 20 weeks | Intervention group n = 50 Psychiatric assessment templates Control group n = 50 Usual paper charts | Median score (max = 100) for adequacy of documentation of 25 items in history, exam, and treatment). | • Significant increase in median score (33 vs 18; p < .01). • Significant increase in proportion of charts documenting education (p = .029), alcohol (p = .045), smoking (p = .009) and interview alone (p = .0001). Non-significant changes for remaining topics. • Overall increase in psychosocial history documentation (9% vs > 1%, p = .003) • Overall increase in newly documented psychosocial problems (16% vs. 10%, p = .05). | 18/27 |
Teo et al. Australia (1995) | PFUP | • Paediatric emergency department • Physicians | Multiple Intervention (education, reminder, and template) 5 weeks | Intervention group n = 52 Phase I: Education and reminders to increase the quality of pediatric asthma documentation (2 weeks). Phase II: Physicians mandated to adopt an acute asthma proforma (3 weeks) Control group n = 204 Usual paper charts. No specific education. | Proportion of charts documenting each of 19 items | • Phase I intervention – education and reminders – resulted in no statistically significant change in documentation. • Phase II intervention – Template/Proforma – significantly improved documentation of 8 of 19 items (p < = .03). | 17/27 |
Van Amstel et al. Canada (2004) | PPN | • Pediatric teaching hospital • Physicians | Reminder/ 4 weeks | Intervention group n = 153 HEADSS (Home, Education, Alcohol, Drugs, Smoking, Sex) stamp in patient charts to remind physicians to document these data items in charts. Control group n = 153 Usual paper charts | • Difference in proportion of charts containing information on psychosocial problems related to: Home, Education, Alcohol, Drugs, Smoking, Sex • Extent of global documentation • Proportion of charts with newly documented psychosocial problems in the above focus areas. | • Significant increase in proportion of charts documenting education (p = .029), alcohol (p = .045), smoking (p = .009) and interview alone (p = .0001). Non-significant changes for remaining topics. • Overall increase in psychosocial history documentation (9% vs > 1%, p = .003) • Overall increase in newly documented psychosocial problems (16% vs. 10%, p = .05). | 20/27 |
Vasileff et al. Australia (2009) | PPN | • Teaching hospital • Emergency department doctors | Facilitation (pharmacist medication verification) 5 weeks | Intervention group n = 29 Pre-admission medication history documented on patients’ charts by pharmacists (and verified by patient’s pharmacy) before patients were seen by emergency department physicians. Control group n = 45 Physicians documented pre-admission medications on a standard form. | • Discrepancies in documented medication histories • Medication errors | • Overall decrease in unintentional medication discrepancy in patients: 3.3% vs. 78.6% (p < 0.05) • Decrease in average number of discrepancies per patient 0.03% vs. 2.51% (p < 0.05). Reduction of missed doses of pre-admission medications 0 vs. 1.04 (p < 0.05) | 13/27 |
Voaklander et al. Canada (2000) | PPC | • Teaching hospital • Emergency department physicians | Multiple Intervention (education and reminder) 13 weeks | Intervention group n = 321 flagged charts; 323 un-flagged charts Injury surveillance training, pocket reminder cards, modification of existing emergency department charts to include chart reminder labels, and space added for inclusion of additional injury related data Control group n = 645 Handwritten unmodified charts | Presence of 14 key data elements included in education intervention | • Significant increase in mean number of 10 of 14 documented data elements - 8.1 flagged charts vs 7.3 unflagged charts vs 6.9 pre-intervention (p < 0.05). • Significant decrease in documentation of prevention measures: (12.1% vs. 21.4%) OR 0.56 (0.38–0.83 CI p > 0.05). • Significant increase (post-intervention) in documentation of activity at time of injury, location of injury, address where injury occurred, adult observer present and environmental conditions (p < 0.05) | 19/27 |
Wrenn et al. USA (1993) | PPN | • Teaching hospital • House staff | Template 35 weeks | Intervention group n = 1129 Structured complaint-specific patient encounter forms for laceration, closed-head injury, pharyngitis, and asthma available to all ER house-staff for 8 months. Control group n = 1276 Usual paper charts | • Proportion of charts documenting 30 aspects of history, physical and treatment • Proportion of charts with complete prescription information | • Mixed results, reported as percentages and odds ratios, across 30 parameters of history taking, physical and treatment – range: 97% vs 17% (OR 176, p < .001) to 98.4% vs 94.4% (OR .28 p: NS) • Significant increase in proportion of charts documenting prescription information (80% vs 73%, p = .007) | 19/27 |
Zick & Olsen USA (2001) | PPN | • Suburban level 1 trauma centre • Physicians | Dictation Not specified | Intervention group n = 47 Dragon Naturally Speaking voice recognition software. 30 min training for physician. Control group n = 47 Traditional voice transcription services | Difference in accuracy (per cent of words correct in document). | Decrease in accuracy of words documented (98.5% vs 99.7% - change of −1.2; CI (−1.5 to −0.8)) | 17/27 |