Introduction
There is a growing awareness that quality and safety failures in healthcare are attributable more to systems and processes than to human errors [
1‐
4]. To address this, healthcare leaders are increasingly applying quality improvement (QI) and process-oriented management practices from other industries including Lean, Six Sigma, Failure Mode Effects Analysis (FMEA), Failure Mode, Effects, Criticality Analysis (FMECA), and operational research and process-oriented costing approaches such as Time-Driven Activity-Based Costing (TDABC) [
5‐
8].
Applying QI methodology is challenging as healthcare processes are highly variable, distributed and multidisciplinary, involving stakeholders with differing interests and motivations [
9‐
15]. Research shows that the success of QI interventions is heavily influenced by their context of implementation [
16‐
18]. Developing interventions that are adapted to the local context and setting is an essential component of successful QI [
19‐
21], along with engagement of stakeholders [
22‐
24]. Process mapping (PM) has the potential to support QI projects in healthcare by engaging stakeholders to create a shared understanding of the systems they are trying to change [
25‐
29]. However, there is little research on the use of PM in healthcare and whether it is achieving its full potential. The term ‘process mapping’ is used to describe several approaches and techniques. Here we refer to the “entire approach that leads to a holistic understanding of the process under review” [
12,
30].
Research shows that the full benefits of PM are accomplished when it is used throughout all the stages of a QI project to plan, implement, monitor and evaluate interventions [
12,
29,
31‐
33]. However, the application of PM within QI initiatives has proved challenging due to the limited time clinicians can devote to it and their limited knowledge of PM methods [
34‐
36].
Although more informed and systematic use of PM in the design and management of healthcare delivery is advocated [
37,
38], there is poor evidence on the use of PM and its effectiveness in healthcare [
39‐
41]. To advance current knowledge on PM and improve its use in practice, we need greater insight into how it works in different contexts, the mechanisms underlying its successful use, and challenges to its implementation [
42,
43].
There is currently no systematic review of the use of PM in healthcare practice. Most published literature only describes empirical studies of individual interventions using PM. There is very limited information on the range and type of healthcare settings in which PM has been used or the benefits of its use. This problem is compounded by the lack of formal criteria to guide the implementation, evaluation and reporting of PM. Some methodological guides focus on PM in healthcare improvement initiatives [
29,
31,
32,
44]; none are based on the structured review of the research evidence. Limited knowledge of the use of PM as a QI method in healthcare hinders its wider adoption [
45]. Therefore, increased awareness of its possible applications and benefits, as well as evidence-based quality criteria for its use, are needed.
This paper reviews the empirical literature and methodological guidance on PM to increase understanding of its use in healthcare to: (1) develop a conceptual framework identifying different phases in PM, with quality criteria for each, to guide the effective implementation, assessment and reporting of this method; (2) identify the context of use of PM in healthcare QI projects; (3) assess adherence of the application of PM as described in empirical literature to the proposed conceptual framework quality criteria and (4) explore the reported evidence for the benefits of using PM in improvement work.
Discussion
The use of PM within healthcare improvement projects helps to support understanding of complex healthcare systems and adaptation of improvement interventions to their local context. We reviewed methodological guidance on PM, peer-reviewed empirical literature, and developed a conceptual framework to guide effective implementation, assessment, and reporting of PM in healthcare. We assessed adherence of 105 empirical studies to quality criteria outlined in a newly created conceptual framework. Comparison of methodological guidelines and empirical literature helped to identify common features characterising the use of PM across the selected studies. We also identified reported context of use and benefits of using PM in improvement work.
To our knowledge, this is the first systematic literature review exploring the use of PM in healthcare improvement projects. The review demonstrates that PM is used in projects to improve quality and safety in a wide range of healthcare settings. These projects focus on different QI tools and approaches, and use PM either as a standalone methodology or as a support for other QI methods.
Using the conceptual framework, we found inconsistencies in reporting and in adherence to PM quality criteria. None of the studies adhered to all the criteria and only 7% studies adhered to 8/10 or 9/10 criteria. Assessment of adherence was, however, challenging due to variation in reporting of PM exercises across studies. This is attributable both to the diversity of the contexts for using PM and lack of standardised reporting requirements. Analysis of the reviewed studies suggests that poor adherence with quality criteria reflects not just problems in the reporting of PM, but also the conduct of the method.
Although for most reviewed studies, views of different stakeholders were gathered, only 15% reported the involvement of those who would be using the processes such as patients/ service users/ customers. Moreover, less than half (45%) clearly reported that process maps were generated through multi-professional meetings. This suggests that some benefits of PM may not have been realised in these studies, as failure to engage all stakeholders is unlikely to produce realistic process maps or support successful patient-centred QI initiatives. If PM is conducted without appropriate stakeholder participation, some of the benefits derived from the social interactions, such as empathy between professional groups and agreement for shared solutions, are inhibited [
12]. Two of the studies identified in the systematic review reported that the limited involvement of clinical staff was related to the difficultly of relieving them from their daily job [
102,
117] but reasons for poor patient involvement should be further investigated [
24].
Only 14 of the reviewed studies report training in PM techniques as part of the project. Limited training in PM techniques may explain the lack of discussion or consideration of the process modelling language used to draw the process map in the reviewed studies. This finding confirms previous research stating that most projects in healthcare only use flowchart diagrams, regardless the variety of process modelling techniques and tools available [
123]. The choice of modelling language used is important in describing and understanding systems analysed with PM and overlooking these aspects can impact its effective use [
124]. Furthermore, training project teams in QI is important not only to improve participants’ technical skills, but also to enhance their engagement in the project [
69,
103].
Some studies reported that they had to balance the rigorous use of the PM method with resource and time constraints they had to face in practice [
48,
93,
103,
113,
117,
125]. Despite reviewed studies demonstrating poor adherence to the identified PM quality criteria, they describe a number of benefits derived from its use in healthcare improvement projects. This demonstrates the key role played by PM in addressing the challenge of designing and implementing change in complex systems. Using PM in improvement work helps to achieve the strategic principles identified by the Successful Healthcare Improvement from Translating Evidence in Complex Systems (SHIFT-evidence) framework (
act scientifically and pragmatically, embrace complexity, engage and empower) [
54]. The capacity of PM to bring together diverse stakeholder perspectives and provide a visual representation of the system is key to address the complexity which characterizes healthcare processes. Within QI projects, PM helped to provide a shared understanding of the reality of complex systems and facilitated dialogue between team members. This increased engagement of project participants and eased their agreement on common solutions to problems, thus supporting two levers recognised as important for successful improvement in complex systems: knowledge co-production and the definition of shared goals across stakeholders [
126,
127].
The use of PM as a monitoring and evaluation tool [
9,
12,
64,
119,
128‐
130] appeared to be out of scope of application by many QI teams. Most of the articles we reviewed focus on use of PM to better understand systems only at the early stages of an improvement initiative or to visualise and disseminate process maps as the “output” of the project. Only 42% of the reviewed studies describe actions undertaken following the PM exercise, suggesting there is still more to know on how PM influences action and impact in overall improvement efforts.
Findings from this literature review show there is still much room for improvement in the use and reporting of PM as a QI method. Limited adherence to recommended practice for PM is a finding consistent with the assessment of fidelity reported for other QI methods [
46,
131].
Implications for practitioners and academics
We unpacked the black box of PM as a QI method and outlined quality criteria to guide its systematic use and reporting. Improving the quality of reporting of PM exercises would enhance transparency, encourage appropriate use of PM in practice, and support the definition of a common language to describe the process of PM [
24]. We encourage practitioners and researchers to use and test the validity of our conceptual framework when implementing or reporting PM. We also suggest further development of reporting guidance for PM exercises and their use as a starting point in the design of prospective studies exploring the effectiveness of the method. Our findings show that improvements in reporting are required not only to systematically describe the “process” of PM but also for representation of the process map, as we found that many articles report only a partial or sample representation of the process map developed. Online versions of published articles or online supplements [
48,
66,
68,
69,
90,
132‐
136] could provide more detailed process maps as these are often difficult to display in printed versions of journals. Improvements in the way process maps are represented and reported might increase the effectiveness of PM as a key QI method. For example, annotating the process map with operational (e.g. waiting times, activity durations, waste/ value), cost (e.g. resources required to perform each activity), patient experience or other project data (e.g. areas targeted or changed by various plan-do-study-act cycles), can be helpful to visually identify gaps in the systems and document the process analysis throughout the project. Previous studies also demonstrated that successful implementation of QI initiatives depends not only on the conformance to methodological guidelines, but is greatly influenced by contextual factors (leadership, organisational culture, etc.) [
16,
137‐
139]. Our study has not taken into account the influence of context on PM exercises, because these factors cannot be assessed by analysis of the literature. While the main contribution of this study is in identifying quality criteria to support a more rigorous use and reporting of PM, we encourage practitioners and researchers to consider the influence of contextual factors on the effective use of QI approaches.
Further research
There is a need for further empirical research to explore the impact of improvement initiative context on practical implementation of PM. We partially explored how PM is used in practice by improvement teams in the NHS in a previous empirical study investigating benefits and success factors of PM in a sample of QI projects [
12]. However, most of the projects included in this study [
12] used the same methodological approach to PM, (multi-stakeholder meetings to generate the process maps). Further empirical research is needed to test whether our findings hold in QI projects developed by teams using different approaches to conduct the PM exercise, as identified in this literature review (e.g. when PM is used within Six Sigma or Lean approaches). Further literature and empirical research could also explore the representation of process maps in more detail. This would provide a wider perspective on how process activities can be represented and annotated with a variety of information (e.g. value/waste, bottlenecks, constraints, patient experience) and how this can influence the effective use of PM within improvement initiatives.
Limitations
There are some limitations due to the search process. The database search could have included other search terms such as “process model*”, “process design*” or “system design*”, but the authors agreed that the effort required to screen the resulting records was not justified by the purpose and boundaries of the present study.
A key limitation is due to the fact that the systematic review is based on PM exercises as described in the selected empirical literature and not on the analysis of actual practice. This implies that results might be affected by reporting bias and selection of studies, as well as publication bias. The content of publications heavily depends on what journals accept for publication and on the limited space allowed. Therefore, projects using specific approaches (e.g. TDABC, Lean or IS development) are less likely to present a detailed description of the PM process, compared to other process improvement projects. Successful projects are more likely to be published than studies reporting less successful interventions, which may be equally useful for knowledge generation. Bias could also arise because we only searched English-language papers. However, our objective was not to perform an exhaustive review of all the studies applying PM techniques in healthcare, nor to assess the effectiveness of PM, but to provide a representative overview of the use of PM as reported in empirical literature.
Another limitation is due to the fact that PM exercises were usually reported as a part of a wider project. Clearly distinguishing the component attributable to PM from that associated with the whole project was therefore not always straightforward. We addressed this limitation in the development of the data item sheet and the conceptual framework, as well as in the data collection and analysis phase. For example, we decided not to quantitatively assess the different roles involved in the PM exercise, because it was not always clear if and how all team members were involved in the PM exercise. Furthermore, we evaluated the actual implementation of the recommendations derived by the PM exercise, considering the improvement actions reported in respect of the whole project.
Finally, within the included studies we found three papers [
120‐
122] which seemed to derive from the same project. We addressed this bias in the analysis and summary phase by discounting the patterns emerging from common characteristics of these three studies.
Conclusions
PM is at the heart of a range of different improvement projects in healthcare. Its effective use is often a fundamental component of successful QI initiatives. If appropriately used, PM brings together perspectives of diverse stakeholders to harness tacit knowledge and understand complex processes, as well as to find common solutions and enhance team engagement. However, variance in reporting and lack of compliance with guiding principles underpinning its effective use may inhibit its full potential in healthcare improvement initiatives, and in sharing learning between initiatives. Greater scientific rigor in the application and reporting of PM is required to increase its effectiveness as a method for improvement and advance the field of improvement science.
The conceptual framework proposed in this paper provides generalisable quality criteria to help “unpack the black box” of PM across a variety of settings and problems in healthcare. We encourage the use and further development of these criteria to guide future adoption of PM and for reporting and evaluating its efficacy. A better understanding of the circumstances surrounding decisions about deployment of mechanisms supporting QI methods, such as PM, is needed in order to increase their effectiveness. Greater recognition of the benefits of PM, as well as training in this method for healthcare professionals and improvement leaders would also contribute to its more extensive and appropriate use in practice.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.