A total of 1002 articles were screened after removing duplicates using titles and abstracts. This resulted in 17 full-text reviews, leading to the inclusion of eight studies in the final analysis (Fig. 1 – PRISMA flowchart).
STUDY CHARACTERISTICS
The analysis included 1797 operative reports, comprising 729 NR and 1068 SR. Table 1 provides detailed information about the included studies. Six studies focused on SR implementation in rectal cancer resections [3, 4, 7, 12, 16, 17], one study on colon cancer resections [2], and one study assessed both types [2]. Except for Bidwell et al. [16], all studies evaluated the completeness of surgical reports as the primary outcome, comparing them to a set of defined variables. The secondary outcomes, such as time to completion, surgeon satisfaction, barriers to implementation, and educational and research value varied across the studies.
Table 1
Summary of included studies and their characteristics.
Author (year) | Procedure | Study Type | Type of narrative report | Type of synoptic report | Narrative reports (number) | Synoptic reports (number) | Primary outcome | Secondary outcome |
Kanters 2020 4 | Rectal cancer resections | Prospective observational | Narrative only report | Any synoptic template included within the report | 48 | 62 | Inclusion of key reporting elements recommended by the National Accreditation Program for Rectal Cancer (same as the ASCRSb rectal cancer checklist) | - |
Bidwell 2020 7 | Rectal cancer resections | Prospective SR data collection compared to retrospectively collected NR. Convergent mixed-methods design. | Narrative report | Synoptic report based on ASCRSb rectal cancer checklist | 180 | 118 | Change in fidelity to documentation of 19 critical items as defined by ASCRSb rectal cancer checklist | Details and perspectives from surgeons about SR |
Buchanan 2022 2 | Colon and rectal cancer resections | Retrospective observational | Narrative report | Synoptic report developed by adapting key colon and rectal cancer-specific quality measures from BCCAa database and consensus between colorectal surgeons | 104 | 491 | Reporting of colon and rectal cancer specific quality measures as developed by adapting key colon and rectal cancer-specific quality measures from BCCAa database and consensus between colorectal surgeons | Synoptic report use |
Maniar 2014 18 | Colon cancer resections | Prospective SR data collection. Retrospective historical NR cohort. | Narrative report | WebSMRc generated synoptic note | 80 | 80 | Reporting of a list of quality indicators extracted from two published articles 25,26, excluding indicators specific to rectal cancer surgery. | Interrater agreement and time to extract data |
Maniar 2015 3 | Rectal cancer surgery | Prospective SR data collection. Retrospective historical NR cohort | Narrative report | Synoptic report developed using the ASCRSb rectal cancer checklist | 97 | 97 | Reporting of a list of quality indicators defined by the ASCRSb rectal cancer checklist | Time to abstract data from reports and inter-rater reliability |
Bidwell 2022 16 | Rectal cancer resections | Prospective SR data collection and compared to a retrospective NR. Prospective mixed-methods study using surveys and qualitative interviews | Narrative report | Synoptic report developed using the ASCRSb rectal cancer checklist, colorectal surgery experts and key stakeholder involvement. 19 items. | 140 | 140 | Acceptability, feasibility, usability. Experiences with implementation and motivations and barriers to use. | |
Edhemovic 2004 12 | Rectal cancer resections | Prospective SR data collection. Retrospective historical NR cohort | Narrative report | Synoptic note developed by the local cancer surgery working group. | 40 | 40 | Completeness of operative reports across set of variables developed by the local cancer surgery working group. | |
Robertston 2020 17 | Rectal cancer resections | Prospective SR data collection. Retrospective historical NR cohort | Narrative report | WebSMRc generated synoptic note | 40 | 40 | Completeness of documentation of quality indicators based on articles by McGory et al 25 and Gagliardi et al 26. | Time to abstract data from the reports. Inter-rater reliability |
a BCCA: Binational Colorectal Cancer Audit |
b ASCRS: American Society of Colon and Rectal Surgeons |
c WebSMR: Web Surgical Medical Record |
INTERVENTION
None of the studies provided a specific definition of NR. All studies, except Kanters et al. [4], used retrospectively collected NR as the comparison group before implementing SR.
Different SR approaches were employed in the included studies. Three studies examining rectal cancer resections developed their SR based on the American Society of Colon and Rectal Surgeons (ASCRS) rectal cancer checklist [3, 7, 16]. Two studies used the Web Surgical Medical Record (WebSMR) program report [17, 18]. Kanters et al. [4] defined SR as an operative note that included any synoptic template within the report. Finally, three studies utilised expert consensus or a surgical working group to develop their SR [2, 12, 16]. Most studies implemented SR using an electronic medical record (EMR), while Bidwell et al. 2022 and 2020 allowed surgeons to dictate synoptic components if an EMR option was not available [7, 16]. Surgeon education before SR implementation was not provided in any study.
COMPLETENESS
Seven studies reported on the completeness of SR compared with NR, assessing various predefined variables. These variables differed among studies, as outlined in Table 1. The mean percentage of included elements detected in SR and NR is summarized in Table 2. All studies showed significantly higher reporting of identified variables in SR compared to NR (P < 0.001). Although a meta-analysis was intended, only one study reported the standard deviation of their data and attempts to obtain missing statistics from authors to impute this data were unsuccessful. Considering the consistent and significant differences across the studies, meta-analysis was deemed unlikely to change the direction of effect.
Table 2
Summary of the data regarding completeness of reporting in NR compared with SR.
Article | NR | SR | NR MEAN % included elements | SR MEAN % included elements | P-value |
Kanters 2020 | 48 | 62 | 39 | 92 | P < 0.001 |
Bidwell 2020 | 180 | 118 | 27 | 95 | P < 0.001 |
Buchanan 2022 | 104 | 491 | - | - | P < 0.001 |
| 74 colon | 347 colon | 43 | 84 | P < 0.001 |
| 30 rectal | 144 rectal | 40 | 84 | P < 0.001 |
Maniar 2014 | 80 | 80 | 24 | 49 | P < 0.001 |
Edhemovic 2004 | 40 | 40 | 46 | 99 | P < 0.001 |
Maniar 2015 | 97 | 97 | 32 | 71 | P < 0.001 |
Robertson 2020 | 40 | 40 | 24 | 54 | P < 0.001 |
TIME TO COMPLETION
Bidwell et al. 2022[16] conducted qualitative surveys to assess surgeons’ opinions on SR implementation. Some surgeons found SR “quick to complete”, whilst others perceived additional documentation as time-consuming, especially when NR were also required [16]. Edhemovic et al.[12] measured the time needed to complete SR. Excluding the learning curve reports, the mean time for completion was 5 minutes 59 seconds. The time for completion of NR was not reported.
SURGEON SATISFACTION
Bidwell et al. (2020 and 2022) explored surgeon satisfaction. Bidwell et al. 2020 found that surgeons believed that SR would help capture critical information in their reports, and they identified multiple uses for SR, including quality assurance, improvement, research, and multidisciplinary team communication [7]. Bidwell et al. 2022 reported high usability, feasibility, and acceptability of implemented SR, including satisfaction across these domains [16].
EDUCATIONAL VALUE
Educational value was mentioned in only one study. Bidwell et al. 2020 found that surgeons perceived SR as a tool that could focus their attention on critical element of the operation, serving as a learning or teaching tool [7].
RESEARCH VALUE
Three studies reported data extraction time and inter-rater agreement for NR and SR.
Maniar et al. 2014 [18] demonstrated significantly faster data extraction with SR compared with NR (2:32min vs 4:01min, P < 0.001) [18]. The interrater agreement was also significantly higher for SR than NR (P < 0.001) [18].
Maniar et al. 2015 demonstrated significantly faster data extraction with SR compared with NR (2:45min vs 4:48min, P < 0.001) [3]. They also found significantly higher inter-rater agreement with SR (P < 0.001) [3].
Roberston et al. [17] reported significantly faster abstraction from SR compared to NR (3:46min vs 6:21 min, P < 0.005).
Bidwell et al. 2022 highlighted that SR implementation facilitated research by enabling easy data collection, accurate data, and faster analysis [16]. Bidwell et al. 2020 found that surgeons believed the data in SR could provide opportunities for quality assurance, quality improvement and contribute to research [7].
BARRIERS TO IMPLEMENTATION
Bidwell et al. 2022 identified several barriers to SR implementation. Surgeons expressed frustration when SR integration into an EMR was not feasible, requiring alternative approaches [16]. Technical and administrative oversight for implementation was lacking in some centres, and additional administrative burden, time pressures and workflow impact were perceived as adverse effects by surgeons [16].