Abstract
Introduction
This systematic review attempts to answer the following question – which strategies to improve clinical nursing documentation have been most effective in the acute hospital setting?
Methods
A keyword search for relevant studies was conducted in CINAHL and Medline in May 2019 and October 2020.
Studies were appraised using the Joanna Briggs Institute (JBI) critical appraisal for quasi-experimental studies. The studies were graded for level of evidence according to GRADE principles.
The data collected in each study were added to a Summary of Data (SOD) spreadsheet. Pre intervention and a post intervention percentage compliance scores were calculated for each study where possible i.e. (mean score/possible total score) × (100/1). A percentage change in compliance for each study was calculated by subtracting the pre intervention score from the post intervention score. The change in compliance score and the post intervention compliance score were both added to the SOD and used as a basis for comparison between the studies. Each study was analyzed thematically in terms of the intervention strategies used. Compliance rates and the interventions used were compared to determine if any strategies were effective in achieving a meaningful improvement in compliance.
Results
Seventy six full text articles were reviewed for this systematic review. Fifty seven of the studies were before and after studies and 66 were conducted in western countries. Publishing dates for the studies ranged from 1991 to 2020.
Eleven studies included documentation audits with personal feedback as one of the strategies used to improve nursing documentation. Ten of these studies achieved a post intervention compliance rate ≥ 70%.
Conclusion
Notwithstanding the limitations of this study, it may be that documentation audit with personal feedback, when combined with other context specific strategies, is a reliable method for gaining meaningful improvements in clinical nursing documentation. The level of evidence is very low and further research is required.
Introduction
Clinical documentation is the process of creating a written or electronic record that describes a patient's history and the care given to a patient (Blair & Smith, 2012; Wilbanks et al., 2016). It serves as an important communication tool for the exchange of information between healthcare providers and it is stored in a printed or electronic medical record (Duclos-Miller, 2016; Mishra et al., 2009). According to Wilbanks et al. (2016) good quality documentation has been defined as documentation that is correct and comprehensive, uses clear terminology, is legible and readable, timely, concise and plausible.
Poor nursing documentation in the acute care setting may have negative impacts on patient outcomes and may also result in litigation (Duclos-Miller, 2016). Therefore it is important to determine if there are any strategies that will provide meaningful improvements in the quality of nursing documentation in the acute care setting.
At the time of writing there were four systematic reviews related to nursing documentation. Three (Johnson et al., 2018; Müller-Staub et al., 2006; Saranto et al., 2014) examined the impacts of standardized nursing languages (SNL) on the quality of nursing documentation. They were narrative reviews, and include studies that were not necessarily confined to the acute sector. They demonstrated that SNL will improve the quality of nursing documentation, assist in the fulfilment of the legal requirements of documentation and facilitate the use of an electronic health record (EHR). One systematic review (McCarthy et al., 2019) examined the effects of electronic nursing documentation and found that utilizing an END system could improve the quality of nursing documentation, decrease documentation errors and increase compliance with nursing documentation guidelines.
These systematic reviews were narrative in structure and no attempt has been made to determine if the improvement in each of the studies reviewed is a clinically meaningful improvement. The aim of this systematic review is to qualitatively and quantitatively analyze the literature in an attempt to determine which strategies to improve compliance with clinical nursing documentation guidelines, and improve the quality of nursing documentation, have been most effective in the acute setting.
Methods
A systematic review of the literature was conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) guidelines, where possible.(Page et al., 2021a, b)
Search Strategy
A keyword search for relevant studies was conducted in CINAHL and Medline in May 2019 and again October 2020, due to the time that had elapsed. The only limitations were for articles in peer reviewed journals that were written in English. An analysis of the text words contained in the titles, abstracts and index terms found in relevant articles was used to inform the search strategy. The reference lists of articles selected for inclusion were hand searched for additional articles. The full search strategy for CINAHL is found in Appendix 1.
Inclusion and exclusion criteria
Studies were included if they were quantitative research investigating strategies to improve clinical nursing documentation in acute hospitals. Where possible, the quantitative components of mixed method studies were also included. The nursing documentation components of studies that also involved allied health or medical documentation were included where possible. Studies were not excluded by intervention, we attempted to include as many studies as possible (See Table 1).
Inclusion and Exclusion Criteria.
Study selection
Abstract and title screening from the database results lists was initially performed by the principal reviewer and citations were downloaded into EndNote X9 if they appeared relevant. The abstracts in EndNote were then screened independently by both reviewers and conflicts were resolved by discussion. Full text screening was undertaken by the principal reviewer.
Data extraction
A Summary of Data (SOD) excel spreadsheet was prepared by the principal reviewer. For each study that met the selection criteria the following data were extracted - author, year of publication, country of origin, study title, aims, study design, setting, sample size, method of randomization, interventions used, instruments used to collect data, statistical analyzes performed, outcome measures, results and conclusions.
Quality appraisal
Studies included in this systematic review were quasi-experimental studies and were appraised for risk of bias by the principal reviewer using the JBI critical appraisal for quasi-experimental studies (Tufanaru et al., 2017). The JBI Critical Appraisal Checklist for Randomized Controlled Trials (Tufanaru et al., 2017) was used for the only RCT included in the review. For before and after studies, the pre intervention group was not considered to be a control group. The statistical analyzes performed in the studies were evaluated for appropriateness with reference to the Flow chart for hypothesis tests, categorical and numerical data, found on the back inside cover of Medical Statistics at a Glance by Petrie and Sabin (2020)
Rating the certainty of the evidence was undertaken using the principles of GRADE when a meta-analysis has not been performed (Murad et al., 2017)
Analysis
The analysis was performed by the principal reviewer. Where possible each study was quantitatively analyzed such that the data collected in each study were used to calculate a pre intervention and a post intervention percentage compliance score i.e. (mean score/possible total score) × (100/1). A percentage change in compliance for each study was calculated by subtracting the pre intervention score from the post intervention score. The change in compliance score and the post intervention compliance score were both added to the SOD excel and used as a basis for comparison between the studies.
For each study, a meaningful compliance rate was defined as a post intervention compliance rate ≥ 70%, using the definition of compliance as defined within the study
Each study was analyzed thematically in terms of the intervention strategies used. The themes were education alone, audit and feedback, EHR versus paper health record, SNL, EHR modifications, new forms, guidelines, and system changes. Each study was coded according to all of the intervention strategies that were applied, and the codes were recorded on the SOD spreadsheet. See Table 2 for a definition of each of the themes.
Definition of the Themes.
This review is as a narrative synthesis with a quantitative component. For each of the thematic strategies, the post intervention compliance scores were compared to determine if any of the strategies were effective in achieving a meaningful improvement in the quality of nursing documentation. Studies that achieved large improvements in compliance from a very low initial compliance base may not have achieved a final compliance rate of ≥ 70%, therefore good performances may have been missed in this analysis. To compensate for this, studies that achieved an improvement of ≥ 50% were also identified and analyzed in terms of strategies employed.
Ethics approval was sought and was not required.
Results
An initial search was performed in May 2019. Due to the time that had elapsed, a follow up search was performed in October 2020. See Figure 1 PRISMA Diagram below

PRISMA diagram for searches conducted in 2019 and again in 2020 due to the time that had elapsed
Studies were excluded after full text review if they did not meet the inclusion criteria for this study (see Table 1).
No studies were excluded after critical appraisal, we attempted to include as many studies as possible.
Seventy six full text articles were reviewed for this study. See Appendix 2 for The Summary of Data table. You can download the data in the SOD spreadsheet from here https://osf.io/8r49s/files/
Fifty seven of the studies were before and after studies, with the remainder being cross sectional studies (six), Plan Do Study Act studies (four), non-randomized controlled studies (four), time course analyzes (two), randomized trials (two) and one randomized controlled trial. Sixty six studies were conducted in western countries including the USA, Canada, Europe, the UK and Australia. The remainder were conducted in Jordan, Kenya, Brazil, Iran and Singapore. Publishing dates for the studies ranged from 1991 to 2020, all but seven of the studies were conducted in the last 20 years.
Nine of the studies included in this review used education as their only strategy to improve nursing compliance with clinical documentation (Cone et al., 1996; Finn, 1997; Griffiths et al., 2007; Jackson, 2010; Lieow et al., 2019; Linch et al., 2017; Müller-Staub et al., 2008; Mykkänen et al., 2012; Phillips et al., 2019). Of these studies, four had a post-intervention compliance rate ≥ 70% (Jackson, 2010; Lieow et al., 2019; Müller-Staub et al., 2008; Mykkänen et al., 2012).
Twenty two studies had audit and feedback as one of the strategies used to improve compliance with nursing documentation (Azzolini et al., 2019; Bernick & Richards, 1994; Cline, 2016; Elliott, 2018; Ellis et al., 2007; Esper & Walker, 2015; Gerdtz et al., 2013; Gloger et al., 2020; Gordon et al., 2008; Goulding et al., 2015; Hayter & Schaper, 2015; Hom et al., 2019; Jacobson et al., 2016; Kamath et al., 2011; Okoyo Nyakiba et al., 2014; O’Connor et al., 2014; Porter, 1990; Stocki et al., 2018; Tejedor et al., 2013; Trad et al., 2019; Unaka et al., 2017; Wissman et al., 2020). Sixteen of these studies had a final compliance rate ≥ 70%. However if the feedback is personal, the number of studies with a compliance rate ≥ 70% improves to ten out of 11 studies (Bernick & Richards, 1994; Cline, 2016; Elliott, 2018; Esper & Walker, 2015; Gloger et al., 2020; Hayter & Schaper, 2015; Jacobson et al., 2016; Kamath et al., 2011; O’Connor et al., 2014; Unaka et al., 2017; Wissman et al., 2020). Download Table 3 Audit and personal feedback from here https://osf.io/8r49s/files/
When audit and feedback are combined with the use of a pre-existing EHR the results are also encouraging (Cline, 2016; Elliott, 2018; Esper & Walker, 2015; Gerdtz et al., 2013; Gloger et al., 2020; Hayter & Schaper, 2015; Hom et al., 2019; Jacobson et al., 2016; Kamath et al., 2011; Tejedor et al., 2013; Unaka et al., 2017). Ten of the 11 studies had a final compliance rate ≥ 70%.
Thirteen studies compared EHR with paper records as one of the strategies for increasing nursing documentation compliance (Akhu-Zaheya et al., 2018; Ammenwerth et al., 2001; Dahlstrom et al., 2011; Gunningberg et al., 2008; Gunningberg et al., 2009; Higuchi et al., 1999; Hübner et al., 2015; Larrabee et al., 2001; Mansfield et al., 2001; Rabelo-Silva et al., 2017; Rykkje, 2009; Thoroddsen et al., 2011; Tubaishat et al., 2015). Six of these studies demonstrated a final compliance rate ≥ 70% when an EHR was utilized. It should be noted that Larrabee et al. (2001) had a high compliance rate, however the improvement was 0.2%. Two studies, (Akhu-Zaheya et al., 2018; Rykkje, 2009) demonstrated a decline in compliance when comparing EHR to paper based records.
Ten studies involved the use of SNL as one of the strategies to improve nursing documentation (Björvell et al., 2002; Darmer et al., 2006; Larrabee et al., 2001; Melo et al., 2019; Müller-Staub et al., 2007; Nøst et al., 2017; Rabelo-Silva et al., 2017; Rykkje, 2009; Thoroddsen et al., 2011; Thoroddsen & Ehnfors, 2007). Four of these studies had a final compliance rate ≥ 70% (Larrabee et al., 2001; Müller-Staub et al., 2007; Thoroddsen et al., 2011; Thoroddsen & Ehnfors, 2007). It should be noted that although Larrabee et al. (2001) had a final compliance rate of 84.2%, the improvement was 0.2%.
Ten studies utilized EHR modifications as one of the strategies to improve nursing documentation (Bruylands et al., 2013; Chineke et al., 2020; Esper & Walker, 2015; Gerdtz et al., 2013; Hom et al., 2019; Jacobson et al., 2016; Kamath et al., 2011; Nielsen et al., 2014; Sandau et al., 2015; Tejedor et al., 2013). Seven of these studies had a final compliance rate ≥ 70% (Chineke et al., 2020; Esper & Walker, 2015; Hom et al., 2019; Jacobson et al., 2016; Kamath et al., 2011; Nielsen et al., 2014; Tejedor et al., 2013). Prompts were used in four of the studies (Chineke et al., 2020; Hom et al., 2019; Kamath et al., 2011; Sandau et al., 2015), and except for Sandau et al. (2015) they all had a final compliance of ≥ 80%.
Thirty two studies used new forms as one of their strategies for improving nursing documentation (Aparanji et al., 2018; Björvell et al., 2002; Bono, 1992; Cahill et al., 2011; Chineke et al., 2020; Christie, 1993; Dahlstrom et al., 2011; de Rond et al., 2000; Dehghan et al., 2015; Elliott et al., 2017; Ellis et al., 2007; Enright et al., 2015; Florin et al., 2005; Förberg et al., 2012; Gerdtz et al., 2013; Gordon et al., 2008; Hayter & Schaper, 2015; Higuchi et al., 1999; Hospodar, 2007; Hübner et al., 2015; Kamath et al., 2011; Karp et al., 2019; Khresheh & Barclay, 2008; Mansfield et al., 2001; Mitchell et al., 2010; Nomura et al., 2018; Nøst et al., 2017; O’Connor et al., 2014; Stewart et al., 2009; Stocki et al., 2018; Thoroddsen & Ehnfors, 2007; Unaka et al., 2017). Seventeen of the 32 studies had a final compliance rate ≥ 70%.
Twelve studies included changes to guidelines, procedures or policies as one of the strategies to improve nursing documentation (Considine et al., 2006; Elliott, 2018; Flores et al., 2020; Gordon et al., 2008; Gunningberg et al., 2008; Habich et al., 2012; Jacobson et al., 2016; Margonary et al., 2017; Mitchell et al., 2010; Nomura et al., 2018; Trad et al., 2019; Turner & Stephenson, 2015). Five of these studies achieved a final compliance rate ≥ 70% (Elliott, 2018; Flores et al., 2020; Gordon et al., 2008; Jacobson et al., 2016; Nomura et al., 2018). It should be noted that the results for Elliott (2018) must be used with caution as some negative results may have left out in the final calculation of compliance rates.
Ten studies included administrative or system changes as one of the strategies to improve nursing documentation (Ammenwerth et al., 2001; Dehghan et al., 2015; Enright et al., 2015; Gerdtz et al., 2013; Kamath et al., 2011; Mansfield et al., 2001; Meyer et al., 2019; Okoyo Nyakiba et al., 2014; Stewart et al., 2009; Trad et al., 2019). All of the system changes improved nursing documentation, six of them had a final compliance rate ≥70% (Ammenwerth et al., 2001; Enright et al., 2015; Kamath et al., 2011; Mansfield et al., 2001; Meyer et al., 2019; Stewart et al., 2009).
Thirty six studies achieved a meaningful compliance rate i.e. a post intervention compliance score ≥ 70%. Download Table 4 Compliance ≥ 70% from here https://osf.io/8r49s/files/
Seven of the studies had an improvement rate of ≥ 50% (Chineke et al., 2020; Gordon et al., 2008; Hayter & Schaper, 2015; Kamath et al., 2011; Müller-Staub et al., 2007; Porter, 1990; Unaka et al., 2017). The post intervention compliance rate was ≥ 80% for all of them, except Hayter and Schaper (2015), where the final compliance rate was 72%.
There are serious concerns regarding the certainty of the evidence, and the evidence has been graded as very low due to methodological limitations and issues with imprecision, inconsistency and publication bias (See Table 5)
Certainty of Evidence.
Discussion
Of the nine studies that used education alone as the strategy to improve nursing documentation, four resulted in a meaningful compliance rate. It was not possible to determine if the form of the education that was applied, e.g. lectures, simulation, case discussion or demonstrations, had any influence on the final outcome of documentation compliance. The number of studies was too small and the descriptions of the education supplied was not always adequate enough to draw any conclusions.
In terms of the hours devoted to education, one study, Müller-Staub et al. (2008), involved 22.5 h of education for a final compliance rate of 94.5%. However another study, Linch et al. (2017) had 30 h of instruction and discussion for a final compliance rate of 45%, therefore it is difficult to draw any conclusions as to whether the amount of time spent on education has any effect on documentation compliance. Bearing in mind the small number of studies that utilized education alone, it appears education alone will improve compliance, however it may not improve compliance to a meaningful extent.
When documentation auditing with personal or individual feedback was one of the strategies utilized, ten out of the 11 studies achieved a compliance rate of 70% or more (see Table 3 https://osf.io/8r49s/files/). This suggests that auditing and personal
feedback, combined with other context specific strategies such as education, new forms, new templates or EHR modifications, may be a reliable strategy for improving compliance. However the results for Elliott (2018) should be used with caution as some negative results may have been left out in the final calculation of compliance rates. When an EHR was used to conduct the audits, the improvement in compliance rate seems to be reliably high; ten out of the 11 studies that used an EHR to conduct the audits achieved a compliance rate of ≥ 70%. This may be due to the fact that the time consuming audit process is made more efficient by using an EHR for data extraction (Lieow et al., 2019).
Thirteen studies utilized an EHR as one of the strategies to improve nursing documentation, six of these achieved a meaningful compliance rate, however for one of these, Larrabee et al. (2001), the improvement was 0.2% It appears from these studies that introducing an EHR may not guarantee a meaningful improvement in nursing documentation, and it may even be counterproductive as shown by the studies Rykkje (2009) and Akhu-Zaheya et al. (2018). In nearly all of the studies the EHR system was not described or named therefore it was not possible to determine if the nature of the EHR had any effect on the results.
Standardized Nursing Language was used in ten studies, four of them achieved a compliance rate of ≥ 70%. The use of NANDA and NIC appears to be more effective than PES or VIPS for improving nursing documentation compliance. Four out of the six studies that used NANDA and NIC had a final compliance rate ≥ 80% (Larrabee et al., 2001; Müller-Staub et al., 2007; Thoroddsen et al., 2011; Thoroddsen & Ehnfors, 2007); however it should be remembered that Larrabee et al. (2001) had 0.2% improvement. Standardized Nursing Language will improve nursing documentation, however it may not improve to a meaningful extent.
Of the 10 studies that used EHR modification, seven achieved a meaningful compliance rate. For the purposes of this systematic review, new or reconfigured EHR templates, are included in the next section - New Forms. Many different EHR modifications were used across the various studies (see Table 2), and there are indications that prompts in the EHR may be useful in improving compliance, however there is no single modification that ensures meaningful compliance.
For the purposes of this systematic review new forms included new or modified paper forms and new or modified EHR templates. Thirty two of the studies included new forms as one of the strategies to improve nursing documentation, seventeen of these studies improved nursing documentation to a meaningful degree.
Of the 12 studies that used the creation of new guidelines or changes to existing guidelines as a strategy, five achieved a meaningful compliance rate. It appears that guidelines will improve nursing documentation but not always to a meaningful degree.
Ten studies utilized administrative or system changes as one of the strategies to improve compliance. The studies were too heterogeneous for any single strategy to be proven effective at providing a meaningful improvement. All of the studies that used system changes showed improved nursing documentation, six out of the ten studies demonstrated improvement to a meaningful degree.
In the future, issues with clinical nursing documentation may be dealt with by technological means, for example the use of digital scribes. Digital scribes will employ advances in speech recognition, natural language processing, artificial intelligence, machine learning and clinical decision support technologies, to translate clinical encounters into meaningful and accurate records (Coiera et al., 2018). However speech recognition systems and artificial intelligence will need to be vastly improved before the benefits outweigh the risks of using a digital scribe. Meanwhile strategies to improve nursing documentation remains a relevant topic for research.
Limitations of the Study
Full text screening and the quantitative and thematic analyzes were performed by one reviewer, this may have caused some bias in the results. There was no funding for this systematic review, therefore full text access to all of the articles that appeared in the search results was not possible, this may have resulted in some relevant research not being included in the final analysis. The research articles that were analyzed in this project were very heterogeneous in terms of design, statistical analyzes and strategies employed, meaning that traditional systematic review meta-analysis was not possible, and the percentage analysis that was undertaken was rudimentary at best. Only data that could be converted to percentages were included, therefore some data were missed in the analysis.
The studies usually combined many strategies, making it difficult to tease out the effects of any single strategy. The time that elapsed between the intervention and the post intervention audits varied between the studies and was not analyzed in this systematic review. This may have disadvantaged the studies that used a longer time interval to determine if the improvements in documentation were sustained over time. The level of certainty of the evidence is very low, (see Table 5). These limitations could be overcome by a follow up study, focusing on audit and personal feedback and utilizing more rigorous statistical analyzes.
Implications for Practice
Documentation audits with personal feedback should be considered as one of the strategies to be utilized when attempts are made to improve the quality of nursing documentation.
Conclusion
Notwithstanding the limitations of this study, it may be that documentation audit with personal feedback, when combined with other context specific strategies, is a reliable method for gaining meaningful improvements in nursing clinical documentation. Utilizing an EHR to perform the audits may be beneficial to the process, by making the audit process more efficient. The certainty in the evidence is very low, therefore using audit and personal feedback as a strategy to improve clinical nursing documentation is an area that would benefit from more research.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
