Article Text
PDF
Abstract
Background Effective handoff practices (ie, mechanisms for transferring information, responsibility and authority) are critical to ensure continuity of care and patient safety.
Objective This study aimed to develop a rating tool (self-rating and external rating) for handoff quality that goes beyond mere information transfer.
Methods The rating tool was piloted during 126 patient handoffs performed in three different clinical settings in a tertiary care hospital: (1) paramedic to emergency room staff, (2) anaesthesia care provider to postanaesthesia care unit (PACU) and (3) PACU nurse to ward nurse.
Results We identified three factors (information transfer, shared understanding, working atmosphere) predicting handoff quality.
Conclusions This study provides insights into the multidimensional concept of handoff quality. Our rating tool is feasible and comprehensive by including not only characteristics of the information process but also aspects of teamwork and, thus, provides an important tool for future research on patient handoff.
- Handoff
- patient safety
- rating tool
- human factors
- teamwork
- care transitions
- assessment
- collaborative
- healthcare quality
Statistics from Altmetric.com
- Handoff
- patient safety
- rating tool
- human factors
- teamwork
- care transitions
- assessment
- collaborative
- healthcare quality
Introduction
Patient handoff as a critical area for patient safety
The work of healthcare teams is distributed over time and location requiring them to employ effective handoff practices to ensure appropriate coordination and continuity of care, especially at care transition points and during shift changes. Patient handoff has been defined as a mechanism for transferring information, primary responsibility and authority that should ideally be a moment of shared cognition or sense-making between healthcare providers1 contributing to informational, relational and management continuity in patient care.2 However, the causes of many adverse events in healthcare can be traced back to inadequate handoffs.3–5
There has been an increasing recognition that a lack of training on teamwork and effective communication combined with a lack of formal handoff systems impede the good practice necessary to maintain high standards of care in all areas of healthcare.6 ,7 Thus, patient handoff is considered a priority for human factors research aiming at improved patient safety.6
Studies on patient handoff
Research on patient handoffs is mostly descriptive. Studies show, for example, that written documentation is rarely used to support care transitions and that interruptions during handoff are frequent.8 ,9 However, little in-depth information is available on teamwork behaviours (eg, shared decision-making) and their effects on handoff quality and safety.
Existing research usually conceptualises handoffs as both episodes of information transfer and as a source of medical error. Consequently, the evaluation of handoff quality has concentrated on accurate and complete information transfer. More recently, additional functions of handoff communication have been pointed out such as training, socialisation and encouraging/maintaining group cohesion, all of which can be related to organisational learning.2 ,10
Few studies have evaluated interventions to improve handoff quality.8 ,11–14 Most interventions involve the implementation of a standardised handoff protocol. Standardisation is a strong force towards a shared understanding of a situation and its demands on a team in the sense of a ‘shared mental model.’15 Studies on standardised handoff protocols in healthcare provide evidence for improved handoff quality. Catchpole and colleagues, for example, found improved handoff quality and teamwork after the implementation of a new handoff protocol without increasing the time needed to perform the handoff.13
A limitation of these studies is that handoff quality was mostly defined as adherence to a handoff protocol, although additional factors such as anticipation, adaptability and adequate strategies for recovery are critical to patient safety in today's complex healthcare environment. Studies in anaesthesia16 ,17 and emergency medicine18 indicate that patient handoffs are a source of error but also provide important ‘audit-points’ essential for prevention, detection and mitigation of failure.1 ,19 ,20 Despite the increasing acknowledgement of this double-edged nature of patient handoffs in the patient safety literature, more research focussing on the potential contribution of patient handoffs to resilience in healthcare is required.21
Methods
Study aim
Our study aimed to develop and test a rating tool for the quality of patient handoff at care transitions that can be used (1) in a variety of clinical settings and (2) for self-assessment by the clinicians involved in the handoff and for assessment by an external observer. Building on conceptual considerations and empirical evidence on patient handoff, we framed patient handoff as an episode of teamwork that involves more than just information transfer.
Our main research question was: what constitutes a safe and effective handoff (ie, which handoff characteristics predict overall handoff quality)?
Instrument development
Critical review of the literature and of existing assessment instruments
In healthcare, as in other high-risk industries, two aspects of performance that are related to safety—often referred to as ‘technical’ versus ‘non-technical’ performance—are distinguished.22 ,23 So far, handoff studies have focused almost exclusively on technical performance (ie, completeness and accuracy of information transfer and adherence to protocols).11 ,14 ,24 ,25 However, it has been pointed out in the literature that patient handoff has important functions that are related to non-technical skills, that is, the cognitive and social skills required in any operational task including decision-making and teamwork.26 This study aimed to integrate both aspects of performance in the assessment of handoff quality.
Most of the previous studies on handoff quality focused on clinicians' general perceptions of and satisfaction with current handoff practice using interviews, focus groups or cross-sectional surveys27–29 or on clinicians' assessment of handoff safety by exploring clinicians' experience with critical incidents associated with poor handoff retrospectively.30 ,31 In contrast, we intended to develop a rating tool for clinicians' and human factors observers' assessment of specific handoff episodes immediately after their completion.32 ,33
In developing our rating tool, we (1) reviewed existing assessment tools for handoff quality33 ,34 and for non-technical performance assessment in healthcare26 ,35 (2) interviewed three experienced healthcare providers focussing on the question ‘what constitutes a safe and effective handoff?’ and (3) conducted initial unstructured field observations. Finally, a prototype of the rating tool was discussed with expert clinicians and industrial psychologists for further refinement.
Rating form
The rating tool used in this study consisted of 16 items describing patient handoff in terms of information transfer and teamwork that had to be rated on a four-point scale (‘agree,’ ‘partially agree,’ ‘partially disagree,’ ‘disagree’) (table 1). Overall handoff quality was assessed using a single item. Because time pressure may influence handoff quality, raters were also asked to rate the time pressure of the clinicians involved in the handoff.
Items of the rating tool for handoff quality
Data collection
The rating tool was used during a total of 126 patient handoffs at care transitions in a tertiary care hospital. To ensure that the instrument can be used in different clinical settings, we included three handoff situations: (1) from paramedics to emergency room staff, (2) from anaesthesia care provider to postanaesthesia care unit nurse (PACU) and (3) from PACU nurse to ward nurse. Research in these settings is particularly important because transitions of care take place in an environment that is event-driven, time-pressured and prone to concurrent distractions while the patient is in an ‘at risk’ state.20
Each patient handoff was assessed independently by three raters: (1) the clinician handing off the patient, (2) the clinician taking on responsibility for the patient and (3) a human factors observer. All assessments were made immediately after completing the handoff.
Data analysis
The two analytical steps carried out in this study investigated (1) the dimensionality and (2) the predictive validity of our rating tool for handoff quality.
Dimensionality of the rating tool: Investigating the conceptual structure of multidimensional constructs using exploratory factor analysis is an important first step in the development measurement tools.36–38 Exploratory factor analysis (principal-component analysis (PCA); varimax rotation) was carried out following the recommendations of Field.39 All ratings of handoff situations in our sample (ie, by three raters in three clinical settings) were included in this analysis.
Predictive validity of the rating tool: Using stepwise regression analysis, we tested the hypothesis that the handoff characteristics included in our rating tool predict clinicians' and human factors observers' perceptions of overall handoff quality in multiple clinical settings (ie, predictive validity of the measurement tool40). In a first step, context variables (ie, rater, setting and time pressure) were entered to control for their potential influence and in a second step, the dimensions of handoff characteristics identified during factor analysis.
Results
Dimensionality of handoff characteristics
A missing values analysis showed that item 4 produced missing answers in about half of the cases, as it could only be rated in case of interruptions. The item was, therefore, excluded from further analyses.
Preanalysis checks showed that the sample size (n=314) is adequate, as indicated by a Kaiser–Meyer–Olkin measure of sampling adequacy (KMO) of 0.81. Further, the Bartlett test of sphericity indicated sufficient intercorrelation between the items (p=0.001). However, by inspecting the intercorrelation matrix, we found that items 14 and 15 correlated perfectly with 1 (ie, raters did not differentiate them) and were thus excluded from further analysis. Indeed, the determinant of the resulting correlation matrix used for subsequent factor analysis was 0.053 (exceeding the required 0.00001), indicating that multicollinearity was no issue.
Exploratory factor analysis (PCA; varimax rotation) revealed three eigenvalues greater than 1, determining the number of factors extracted together with the loading pattern (table 2). The three factors account for 49.96% of the variance in the items.
Three factors of handoff characteristics (rotated factor loadings higher than 0.3)
Factor 1 ‘Information transfer’ deals with the ‘technical aspects’ of patient handoff: (1) the transmission of the clinical information (ie, selecting and communicating all relevant information, handing over a complete documentation, communicating an assessment of the patient, addressing priorities for further treatment) and (2) the organisation of the handoff that may support effective information transfer between clinicians (ie, continuously using available documentation, allowing for enough time and following a logical structure).
Factor 2 ‘Shared understanding’ deals with the means for establishing a shared understanding between the clinician handing off the patient and the clinician taking on responsibility for the patient. It concerns closing the communication loop (ie, ensuring completeness of the handoff, actively clarifying questions and ambiguities) and discussing possible risks and complications in addition to the routinely communicated information that falls into factor 1.
Factor 3 ‘Working atmosphere’ deals with the atmosphere within the team of clinicians involved in a handoff (ie, tensions, establishing good contact) and with respect for the patient (ie, considering the patient's experience).
Factors predicting handoff quality assessments
Overall, the ratings of handoff quality and the three factors identified in this study were rather positive (table 3).
Median values and IQR for ratings of handoff quality and the three factors
Correlational and multiple regression analysis showed that all three factors identified in this study had good predictive validity for the perceived overall handoff quality. Information transfer showed the highest correlation with perceived handoff quality (r=0.54, p≤0.001), followed by shared understanding (r=0.40, p≤0.001) and working atmosphere (r=0.19, p≤0.01).
Including the context variables in the first step of a stepwise regression analysis revealed essentially the same relationships between the three factors and perceived handoff quality: information transfer (β=0.59, p≤0.001), shared understanding (β=0.28, p≤0.001) and working atmosphere (β=0.16, p≤0.01).
Discussion
We have developed and tested a comprehensive rating tool for handoff quality that goes beyond previous research which has focused mainly on information transfer in combination with handoff protocols.
This study provides valuable insights into the complex, multidimensional concept of patient handoff quality by identifying three general themes: (1) information transfer, (2) shared understanding and (3) working atmosphere. These themes, which may be considered key issues to ensure safe and effective patient handoff at care transitions, were identified across observers with clinical and human factors background and across three clinical settings. They may, therefore, serve as a starting-point for fine-grained analysis and for an expansion to other handoff situations. A general conceptual model of handoff quality will allow for comparing different samples (eg, homogeneous vs heterogeneous professional groups) and, thus, will help to identify targeted interventions for a variety of handoff situations.
Limitations
Although we included three different clinical settings, the handoff situations studied share key characteristics (for classifications of handoff situations see Behara et al and Cheung et al).9 ,41 All handoffs involved transferring and receiving clinicians with different professional background and a single patient that was in most cases novel to the receiving clinician. Thus, the dimensional structure found in this study may not generalise to handoffs of multiple patients between homogeneous groups of clinicians (eg, during shift changes).
It may be because of this restriction to specific types of handoff situations in a single hospital that the distribution of answers did not allow for a statistical analysis of the contribution of (1) interruptions (item 4) and (2) imbalance between completeness and information overload (items 14 and 15) to handoff quality. Studies including a broader spectrum of handoff situations seem warranted to further explore these issues. However, the result that interruptions were present in about 50% of all handoffs observed in this study—leading to the exclusion of this item from the methodological analysis of our rating tool—underlines their relevance, which is frequently discussed in the patient safety literature.8 ,9
Future studies in this area may have to use an adapted version of the rating tool to capture relevant aspects of other handoff situations and, given the recent efforts to improve handoff practices, should also include preventive measures taken by clinicians in order to improve handoff quality (eg, attempts to minimise interruptions). Although implicitly covered by the ‘shared understanding’ dimension of our rating tool, an area for further development may be to explicitly account for an important but often overlooked aspect of patient handoffs: potential for recovery from failure by auditing the previous work and decisions.1 ,16–20
Because we used only subjective assessments of handoff quality, our conclusions must be considered preliminary. The study relies on self-report data for independent and dependent variables and, thus, factors such as common method variance or social desirability may have influenced the correlations in this study in unknown ways. Additionally, we assessed handoff quality immediately after completion of each handoff. However, future studies need to include additional measures to allow for comparisons of handoff characteristics and objective outcomes.
Conclusion
In summary, this study has implications for understanding the complex nature of handoff quality and how it can be effectively measured. This presents an important methodological step regarding the systematic investigation of factors affecting handoff quality in a multitude of healthcare settings and should ideally be complemented by detailed analyses of the interaction processes during patient handoff.42 It can also inform interventions aiming at improved patient safety. For example, there is an increasing concern in human factors research that detailed procedural rules do not necessarily support human performance, especially in critical situations where strong, but flexible guidance is required.43 Thus, standards for safe and efficient patient handoff should go beyond structured information transfer and include elements of teamwork allowing for adaptation and flexibility.21 ,41 ,44 In order to support healthcare professionals adequately, experiences from other industries need to be complemented by detailed studies identifying successful handoff practices for different healthcare settings and integrate this knowledge in process improvements and educational efforts.
References
Footnotes
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.