Research on nursing handoffs for medical and surgical settings: an integrative review
Abstract
Aims
To synthesize outcomes from research on handoffs to guide future computerization of the process on medical and surgical units.
Background
Handoffs can create important information gaps, omissions and errors in patient care. Authors call for the computerization of handoffs; however, a synthesis of the literature is not yet available that might guide computerization.
Data sources
PubMed, CINAHL, Cochrane, PsycINFO, Scopus and a handoff database from Cohen and Hilligoss.
Design
Integrative literature review.
Review methods
This integrative review included studies from 1980–March 2011 in peer-reviewed journals. Exclusions were studies outside medical and surgical units, handoff education and nurses' perceptions.
Results
The search strategy yielded a total of 247 references; 81 were retrieved, read and rated for relevance and research quality. A set of 30 articles met relevance criteria.
Conclusion
Studies about handoff functions and rituals are saturated topics. Verbal handoffs serve important functions beyond information transfer and should be retained. Greater consideration is needed on analysing handoffs from a patient-centred perspective. Handoff methods should be highly tailored to nurses and their contextual needs. The current preference for bedside handoffs is not supported by available evidence. The specific handoff structure for all units may be less important than having a structure for contextually based handoffs. Research on pertinent information content for contextually based handoffs is an urgent need. Without it, handoff computerization is not likely to be successful. Researchers need to use more sophisticated experimental research designs, control for individual and unit differences and improve sampling frames.
What is already known about this topic .
- Nursing handoffs can create gaps, errors and omissions in care.
- Handoffs are considered complex activities, having been a topic of research since 1969.
- The majority of previous authors published handoff studies about nurses' perceptions using instruments without established psychometrics.
What this paper adds
- Handoffs meet the criteria to be considered a ritual and serve multiple, important functions beyond information transfer.
- Prescribing one handoff method across units ignores specific unit and patient characteristics. Determining a standard and method for handoffs may be less important than creating a consistent, predictable structure and method tailored to individual medical and surgical units.
- An urgent need exists to determine contextually based, pertinent information for handoffs on medical and surgical units.
Implications for practice and/or policy
- The current adoption of beside handoffs on medical and surgical units is not supported by research findings nor is it suitable for handoffs for every unit, e.g. where sensitive information is exchanged (palliative care, oncology, psychiatry), or on units with patients prone to infections such as transplant units.
- Future work should include how face-to-face handoffs provide support for error-detection and dispelling erroneous assumptions.
- Without knowing pertinent contextual information tailored to medical and surgical units, practitioners are unlikely to be successful at computerizing the process.
Introduction
Patient care handoffs, or change of shift reports, can create important information gaps, omissions, errors and patient harm (WHO 2007, Kitch et al. 2008). Authors found inadequate handoffs were related to sentinel events (Croteau 2005), critical incidences (Pezzolesi et al. 2010), errors and near misses in novice nurses (Ebright et al. 2004) and omissions requiring information searches before nurses could begin patient care (Lange 1992).
Worldwide, nursing handoffs are a topic of continuing interest. Unfortunately, the majority of published material is anecdotal or concerns nurses' perceptions about the merits of various methods for handoffs. These reports, although interesting, do not assist readers in critically analysing findings because published reports described positive results across all handoff methods and analyses.
At least eight reviews on nursing handoffs are available (See supporting information Table S1 in the online version of the article in Wiley Online Library), (Hays 2003, Strople & Ottani 2006, Arora et al. 2009, Messam & Pettifer 2009, Cohen & Hilligoss 2010, Patterson & Wears 2010, Riesenberg et al. 2010, Matic et al. 2011). These reviews add to our knowledge but gaps in the research remain. Authors have compiled results across nursing in distinctly different contexts despite differences in work design, information needs and technology support requirements. Two studies combined results across professions (Arora et al. 2009, Cohen & Hilligoss 2010). Findings across different contexts (settings) and professions can be difficult to synthesize and generalize. Authors have called for technology support for handoffs (Strople & Ottani 2006, Matic et al. 2011). One set of authors, Riesenberg et al. (2010), limited studies to the USA, eliminating relevant international research and combining results across all settings. They also included studies about nurses' perceptions; however, these studies were based upon instruments without established psychometrics. Another included informal work in a review (Strople & Ottani 2006). No synthesized information exists on research outcomes for nursing handoffs on medical surgical units. This integrative review was completed to address that gap. A research focus on specific contexts (e.g. medical surgical) is then paramount because understanding contextually based handoff information is a prerequisite for successful computerization (Messam & Pettifer 2009, Staggers et al. 2011).
Various terms are used for this communication event: handover, handoff, signout, signoff, intershift report, change of shift report and shift report. The authors use the term ‘handoff’ because of its prevalence in the literature; it is defined as ‘the exchange between health professionals of information about a patient accompanying either a transfer of control over, or of responsibility for the patient’ as proposed by Cohen and Hilligoss (2010, p. 494).
The baseline purpose of nursing handoffs is information transfer, but handoffs provide other important, implicit functions across settings: educational (Kerr 2002), emotional (Parker et al. 1992), group cohesion (Parker et al. 1992, Lally 1999), communicating group values (Lally), accountability/transfer of responsibility (Patterson & Wears 2010), social (Kerr 2002, Patterson & Wears 2010), a chance to demonstrate clinical prowess (Wolf 1989), transparency to identify any erroneous thinking (Patterson & Wears 2010), socializing novice nurses into the expected role of the nurse (Wolf 1988, 1989, Lally 1999) and keeping nurses' anxiety at bay on units with dying patients (Evans et al. 2008). Parker et al. (1992) stated more broadly that handoffs were important in defining a professional identity for nursing. Moreover, 20–30% of information shared in handoffs is not documented in the health record (Sexton et al. 2004) including nurses' clinical judgments about patients (Lamond 2000).
Nursing handoffs have been a topic of interest since at least 1969 (Clair & Trussell 1969). A majority of authors concentrated efforts on practical aspects of handoffs, describing the merits of the myriad of handoff methods including: written (Baldwin & Mcginnis 1994); phoned including S-BAR or Situation, Background, Assessment and Recommendation (Brown 2007); taped (Prouse 1995); care-plan based (Clemow 2006); bedside (Monahan et al. 1988, Parker et al. 1992, Webster 1999); and reading the patient's chart at the bedside (Tucker et al. 2009). Nurses had positive perceptions about these various types of handoffs on effectiveness and efficiencies for these measures: accuracy (Monahan et al. 1988), quality (Prouse 1995), relevancy (Prouse), conciseness (Prouse), completeness (Monahan et al. 1988, Tucker et al. 2009), improved documentation (Clemow 2006), satisfaction with the method (Mckenna & Walsh 1997, Anderson & Mangino 2006, Clemow 2006), perceived effectiveness (Webster 1999), perceived safety (Chaboyer et al. 2010), teamwork (Chaboyer et al. 2010), confidentiality (Webster 1999), ability to care for patients (Sidlow & Katz-Sidlow 2006) and access to information (Webster 1999). Authors reported decreased overtime with bedside report (Monahan et al. 1988, Anderson & Mangino 2006), taped report (Prouse), phone report (Brown 2007) and increased efficiency and teamwork (Chaboyer et al. 2010).
The disparate findings from these less formal studies are not only difficult to reconcile but may provide spurious information to readers. Perceptual study results are largely based upon survey instruments without established psychometrics (Riesenberg et al. 2010). However, readers may and reviewers have (Hays 2003, Strople & Ottani 2006, Messam & Pettifer 2009, Riesenberg et al. 2010, Matic et al. 2011) weighed findings from these less formal reports equally with the findings from more formal ones.
A fundamental assumption of this time-honoured event is that information during handoffs must be accurate and complete enough to allow nurses to plan for care efficiently and effectively. Handoffs are assumed to contribute to the quality and continuity of care; without an adequate handoff, nurses' care of patients would be compromised (Dowding 2001). Interdisciplinary literature addresses the topic. Authors examining 10 years of interdisciplinary sentinel events found communication breakdowns, including handoffs, were implicated in two-thirds of the events (Croteau 2005). A retrospective review of 36 months of critical incidents identified 334 handoff incidents. The most common issues were incomplete handoffs (45%) or no handoff at all (29%) (Pezzolesi et al. 2010). Inadequate nursing handoffs, specifically the lack of information or confusing information, were involved in seven of eight cases of novice nurses' errors or near-misses (Ebright et al. 2004) and research showed that nurses had to retrieve additional information before they could begin patient care activities (Lange 1992). Task uncertainty in different contexts lengthened handoff duration (Mayor et al. 2011). In a review of surgical malpractice claims, 60 cases involved 81 communication breakdowns; 43% were with handoffs (Greenberg et al. 2007). Authors (Nagpal et al. 2010b) found that only 43·9% of key information was retained by the time patients were transferred from the operating room to the gastroenterology inpatient unit and 75% had incidences. Only Patterson and Wears (2010) defined handoffs as a form of resilience where erroneous assumptions, actions and error-detection are made transparent. Although this brief examination of existing literature might conclusively indicate the issues in handoffs, the majority address physician handoffs. The literature is not yet synthesized for handoff outcomes on medical and surgical units.
The review
Aim
This integrative review was in preparation for computerizing handoffs. The purpose was to synthesize the outcomes of nursing handoff research on medical and surgical units.
Design
An integrative review was chosen to synthesize study methods and findings. Integrative review is a specific method to summarize past literature and to give a comprehensive understanding of an area (Whittemore & Knafl 2005). This technique is useful to outline the state of the science, contribute to theory development and it also has applicability to practice and policy. Unlike systematic reviews that include only quantitative studies, this review method includes all study designs while still adhering to rigorous review processes. Because of difficulties described above, studies about nurses' perceptions and informal reports were excluded from this review.
Search methods
The search strategy included literature from 1980–March 2011 from PubMed, CINAHL, Cochrane, PsycINFO, an extensive list of hospital-based handoff research articles compiled by Cohen and Hilligoss, and Scopus databases. The search terms were nurs$ AND research AND handoff$ OR handover$ OR intershift report$ OR shift report$ NOT chemistry or biochemistry. The latter NOT terms were included because the ‘shift’ term yielded results about chemical shifts. Relevant articles were those with reported research on nursing handoffs, papers published in English in peer-reviewed journals, research completed on medical and/or surgical units, using any one of the methods, and including any one of the variables except those noted below. Studies were included if they ranged from an admission point such as the emergency department to medical and surgical units. Any study design was deemed relevant and any method whether qualitative, descriptive or experimental. Studies were excluded if they were focused entirely on non-medical and surgical settings such as emergency departments, critical care or perioperative areas, discharges off the unit, interfacility transfers, long-term care and studies with outcome variables on nurses' satisfaction, attitudes, or perceptions about handoffs. Articles about handoff education and conference proceedings were excluded, the latter because review processes can be highly variable and this review required articles having undergone the peer-review process for established journals.
Search outcome
The search strategy yielded a total of 247 references, 62 of which were duplicates (Figure 1). All citations were rated for relevance using a scale of relevant, not relevant or of questionable relevance by both authors. Articles classified as not relevant were eliminated because they focused on a setting or profession not of interest, e.g. critical care or physicians. A set of 81 was retrieved for further analysis. All 81 were read because of the difficulty in determining relevance by examining only the abstract and article title. Reviews were noted and set aside. References from articles were assessed for inclusion by the authors. Subsequently, all retrieved articles went through another round of relevance rating, culminating in a final set of 30 research studies.
Data abstraction
The primary author initially abstracted data and the second author verified/edited data.
Quality appraisal and synthesis
Studies were first divided into groups according to study design (experimental, descriptive and qualitative) and independently evaluated by the primary author and two research assistants using a quality appraisal checklist developed by Bowling (2002) and modified by Desborough et al. (2012). Discussion resolved the discrepancies for 100% consensus. Appropriate for an integrative review, this appraisal included clarity in aims, methods, statistics, discussion and implications (See supporting information Table S2 in the online version of the article in Wiley Online Library). However, the Bowling checklist did not address possible threats to validity for experimental studies; therefore, these were also evaluated using a tool titled QUality ASsessment Informatics Instrument or QUASII (Weir et al. 2009). The tool was developed using the threats to validity taxonomy identified by Shadish et al. (2002). Face validity was established by expert review; inter-rater reliability ranged from 0·85–0·94. In this study, the QUASII was modified to remove informatics-specific items, i.e. ‘To what degree was the implementation of the information system sufficient?’ The rating scale for each of the 12 items was modified from a Likert scale of 7 points to 3 to indicate low, medium and high quality scores because discriminating among 7 points was highly subjective. Discriminating quality differences between a score of 4 or 5 was difficult and subtle. The authors concluded that a reduced number of categories (low, medium and high) with clear definitions would allow better classification. Better quality determinations and more consistent rationale were then possible. The total number of possible points for this instrument is 36. The authors completed QUASII ratings; inter-rater reliability was 0·91. The three studies were eliminated from further consideration because they achieved a QUASII score below 10. Likewise, their Bowling checklist score was low.
Methods in Whittemore and Knafl (2005) were used to synthesize studies including a constant comparison method. Data were extracted item by item according to pre determined fields relevant to handoffs (Table 1; See supporting information Table S3 in the online version of the article in Wiley Online Library). The overall classification scheme was based upon the study purpose and inductively derived after careful scrutiny of study outcomes. The themes are: information effectiveness and efficiency, handoff processes and structures, handoffs as rituals and information content.
| Source | Purpose, setting, sample, handoff method, care model | Design and methods | Major findings |
|---|---|---|---|
| Handoff effectiveness and efficiency | |||
| Berkenstadt et al. (2008) | Purpose: Improve communication of key information. Setting: Simulation, hospital, Israel. Sample: 25 staff, 3 patients; Handoff method: Verbal; Care model: NA | Design: pre-post test. Methods: 224 pre- 166 post (6–8 week lag), multiple methods | Increased key information, e.g. physiologic data, goals, IV pumps. No change in checking monitor alarms, ventilator |
| Dowding (2001) | Purpose: Compare handoff style and content on recall, care planning. Setting: Laboratory, 2 hospitals in Scotland. Sample: Convenience sample, 48 nurses; Method: two styles; Care model: NA | Design: Experimental 2 × 2 factorial. Methods: Random assignment, 8 scenarios; ‘gold standard’ care plans, scored checklist for accuracy | Low accuracy recall rate. Retrospective style higher for care planning, not recall, accuracy; consistent style higher for recall, accuracy, not planning. Verbal not best |
| Fenton (2006) | Purpose: Improve handoff quality, completeness. Setting: 26-bed unit for older adults, rehabilitation hospital, UK. Sample: Random selection of 15 patients from 5 handoffs (convenience). Handoff method, Care Model: Not discussed | Design: Pre-post. Methods: Audit pre- and post implementation (6 months). ‘Essence of Care’ compared to documentation. No statistical analyses | Medical information more frequent than nursing care; Significant omissions. Post = improvement in 13 categories. No change in time |
| Mckenna and Walsh (1997) | Purpose: Handoff redesign, evaluate efficiency, effectiveness. Setting: High dependency unit, oncology/palliative care, hospital, Australia. Sample: 14-day trial. Handoff method: Various; Care model: Not discussed | Design: Descriptive. Methods: Unit developed, implemented a model (bedside) or combination (verbal, bedside); audits post implementation only | All reduced times to 30 minutes or less. Bedside report difficult on palliative care due to sensitive information |
|
Nagpal et al. (2010a) Study #1 |
Purpose: Evaluate tool on information completeness. Setting: Surgical service, teaching hospital, UK; Sample: Convenience, 20 patients, 18 physicians, nurses. Handoff method: Unknown; Care model: Not discussed | Design: Descriptive. Methods: Interviews, observations, records review. Patients traced from pre-op to unit | Information transfer failures in each phase; 43·9% essential information reached surgical unit. Adverse events in 15 patients. Information redundancy is crucial |
| Nelson and Massey (2010) | Purpose: Develop, test, implement a standardized electronic template. Setting: 32-bed surgical oncology unit, US; Sample, Handoff method, Care model: Not discussed | Design: Pre-post Quality Improvement project. Methods: Microsoft Excel template with no EHR | Average time in report decreased by 38 minutes after 6 and 11 months, savings of $73,726 in overtime costs |
| O'Connell et al. (2008) | Purpose: Compare 3 types of handoffs on effectiveness. Setting: 5 units, teaching hospital, Australia; Sample: Convenience, 40 nurses. Handoff method: Bedside, taped, verbal; Care model: Not discussed | Design: Qualitative, grounded theory. Methods: Interviews, observations. Grounded theory, used NUD*IST software; interview guide not available | Lists strengths, limitations of each method. Information varied by nurse, influenced by contextual, patient, nurse factors. No one type is most effective |
| Pothier et al. (2005) | Purpose: Compare 3 handoff styles on data loss. Setting: Laboratory setting, St. Michael's hospital, UK; Sample: Convenience sample, 5 nurses, 12 simulated patients (same 21 data points) over 5 cycles. Method: Verbal; Care model: NA | Design: Quasi-experimental; Methods: Random assignment to style, videotaped, waited 60 minutes before next handoff, 3 investigators rated accuracy | Verbal – loss of data after 3 cycles; traditional note-taking (written) – 31% accuracy at 5 cycles; typed sheet with verbal – minimal loss. Use formal handoff sheet, verbal |
| Richard (1988) | Purpose: Compare information in handoffs to patients' actual conditions. Setting: 19 units, 800-bed hospital, Western, US. Sample: Convenience, 57 handoffs for 584 patients. Method: Not discussed; Care model: Primary care | Design: Descriptive, exploratory. Methods: Listened to handoffs, visited, assessed patients immediately. Data collection sheet – 11 items | Incongruence in every handoff (most common = IV sites). Taped v. verbal = more omissions, incongruence |
| Sexton et al. (2004) | Purpose: Assess handoff content, formal documentation. Setting: 30-bed medical unit, 1 hospital, Australia; Sample: 23 audiotaped. Handoff method: Nurse in charge, verbal handoff to all staff in office; Care model: Not discussed | Design: Qualitative, descriptive. Methods: Content analysis of audiotaped handoffs. counted information categories | Style, duration, structure, content highly variable. 93·5% information already in documentation, 5·9% important, not available in documentation. Use written. |
| Handoff processes, structures | Design and methods | Major findings | |
| Benham-Hutchins and Effken (2010) | Purpose: Identify patterns of communication from the Emergency Dept to inpatient unit; Setting: University hospital, U.S., 4 units. Sample: Convenience, snowball 5 handoffs, multiple professions. Handoff method, Care model: Not discussed | Design: Descriptive, exploratory. Methods: Observation, social network analysis | Network size 11–20. Exchanges non-linear, complex, unpredictable, shaped by providers' information needs. Communicate across unit and professional boundaries |
| Chaboyer et al. (2010) | Purpose: Describe structure, processes, perceptions of bedside handoffs; Setting: 6 units, 2 regional hospitals, Australia. Sample: Convenience, 532 handoffs. Handoff method: Various including bedside; Care model: Team nursing | Design: Qualitative, case study. Methods: Semi-structured observations | Patients participated in less than 1/2 of handoffs, 1 hospital used SBAR, report only on assigned patients, printed handoff sheet had primary role during, after |
| Engesmo and Tjora (2006) | Purpose: Compare verbal to reading EHR report. Setting: 2 units, one with full EHR, large hospital, Norway. Sample: 2 case studies, Handoff method: Verbal, electronic; Care model: Primary, team nursing, verbal handoffs | Design: Qualitative, case study. Methods: Pre-had verbal reports; post semi-structured interviews, observations, emails, documents | Replacing face-to-face communication with electronic patient systems may hinder nursing handoffs |
| Hopkinson (2002) | Purpose: See how handoffs supported nurses caring for dying patients. Setting: 8 medical units, 2 hospitals, UK. Sample: 28 nurses, 10 bedside, 18 office handoffs; Handoff method: Verbal, office, bedside; Care model: Not discussed | Design: Qualitative, phenomenological. Methods: Semi-structured interviews, pattern identification | Handoffs provide emotional support for nurses caring for dying patients; provide information for nursing decisions, actions |
| James et al. (2010) | Purpose: Knowledge construction in nursing practice. Setting: 26-bed, surgical university hospital, Sweden; Sample: Convenience, 14 nurses, 11 assistants. Method: Read documentation, asked questions; Care model: Worked in pairs | Design: Qualitative, ethnographic, hermeneutic approaches. Methods: Participant observation, interviews, reviewed documents, thematic analysis | 9 themes. Mainly practical knowledge, teche, episteme. Knowledge construction moves between known, unknown driven by suspicion, interaction |
| Kerr (2002) | Purpose: Better understanding of functions, practices of handoffs. Setting: 2 units, hospital, UK. Sample: Convenience sample, 20 handoffs, shadowed 1–2 nurses. Handoff method: taped, team; Care model: Not discussed | Design: Qualitative. Methods: Observation, semi-structured interviews, field notes. Quantitative overview, thematic analysis | Complex, multiple functions. Knowledge, expertise hidden in handoffs. Accomplishes informational, social, educational functions |
| Lally (1999) | Purpose: Investigate functions of nurses' communication, social processes. Setting: 1 surgical unit, general hospital, UK. Sample: Convenience, 6 handoffs. Handoff method: Not discussed; Care model: Team nursing | Design: Qualitative, ethnographic approach. Methods: Unstructured observations, thematic analysis | Transfer of patient information, nursing goals, values, team building, Information categories: nursing process, model in action, them and us, learning the ropes |
| Payne et al. (2000) | Purpose: Clinical discourses in acute elderly units. Setting: 5 units, general hospital, UK. Sample: Convenience, handoffs, interviews. Handoff method: verbal handoff for all staff, then teams. Care model: Teams with primary nurses | Design: Qualitative, ethnographic approach. Methods: Observation, interviews, document review. Used Ethnograph, grounded theory | Formulaic, cryptic, high speed, abbreviations, required socialized knowledge to interpret. Prioritized biomedical, physical. Inadequacy still allowed adequate care |
| Sherlock (1995) | Purpose: Nature of unit handoffs. Setting: 2 medical units, hospital, UK. Sample: Convenience, 28 nurses, 8 student nurses. Handoff method: Not discussed; Care model: Team nursing | Design: Qualitative, participant observation. Methods: Observations, field notes, focused, unstructured interviews over 2 weeks | Lasted 10–60 minutes by senior nurse, complex process, very comprehensive, variable quality, no standardization |
| Handoffs as a ritual | Design and methods | Major findings | |
| Ekman and Segesten (1995) | Purpose: Identify, describe, explain rituals. Setting: 23-bed internal medicine unit, university hospital, Sweden. Sample: Convenience, afternoon reports × 1 month, 1 unit. Handoff method: Verbal; Care model: Team and ‘task allocation’ | Design: Qualitative, ethnographic approach. Methods: Participant observation, field notes. Thematic analysis with member checks | Range of rituals, main = deputed power of medical control. Hidden functions such as socialization. Minimal attention to nursing care, mostly medical care |
| Holland (1993) | Purpose: Investigate nursing culture. Setting: 1 unit (unspecified), hospital, UK. Sample: 3 key informants. Method, Care model: Not discussed | Design: Qualitative, ethnographic approach. Methods: Participant observation, interviews | Rituals exist, not harmful. Handoff us a ritual with common values, symbolic language ensuring exclusivity |
| Evans et al. (2008) | Purpose: Evaluate handoff as a ritual, how anxiety served as a function to organize practice. Setting: Hospital (units not specified), Australia. Sample: Convenience, 14 handoffs, 1 week apart. Method, Care model: Not discussed | Design: Case study, qualitative. Methods: Discourse analysis of field notes | Handoffs met criteria for rituals, alleviated anxiety. Had stereotyped comments, used a concluding remark, conscious non-interrupted attendance, unwritten laws |
| Strange (1996) | Purpose: Examine function, meaning of handoffs. Setting: 1 unspecified unit, noted), hospital, UK. Sample: Unspecified observation. Handoff method: Charge nurse on all patients' conditions, then dyad. Care model: primary nursing | Design: Qualitative, ethnographic approach. Methods: Participant observation. Observations made while the author worked as a team member | Handoff has technical function, psychological, social protective functions, values/rules, rigid hierarchy |
| Wolf (1988) | Purpose: Describe nursing rituals. Setting: 32-bed medical unit, teaching hospital in the US. Sample: Convenience, 6 RNs, 9 graduate nurses. Handoff method: verbal; Care model: Not discussed | Design: Qualitative, ethnographic. Methods: Participant observation, field notes, audio taped, interviews, document, event analysis | Nursing rituals include change of shift report. Serves as socialization, accountability, responsibility for patient care, use of nursing-specific language |
| Wolf (1989) | Purpose: Describe the language of handoffs. Setting: 32-bed medical unit, teaching hospital, US. Sample: 29 staff with 6 RNs, 9 graduate nurses. Handoff method: verbal; Care model: Not discussed | Design: Qualitative, ethnography. Methods: Participant observation, field notes, interviews, audio taped, document, event analysis | Nursing language, jargon serves as professional socialization, complex meanings in shorthand, skill with jargon reflects clinical ability, excludes non-nurses |
| Information content, tools | Design and methods | Major findings | |
| Hardey et al. (2000) | Purpose: Communication for patient care, processes, structure. Setting: 5 units for elderly care, general hospital, UK. Sample: 23 handoffs, 24 interviews. Handoff method: verbal in office, at bedside. Care model: Primary nursing | Design: Qualitative, ethnographic approach. Methods: Observation, field notes, audio taping, interviews, document review | ‘Scraps’ may have an important role in information communication, delivery of care. Processes for constructing scraps, content & role, confidential nature |
| Lamond (2000) | Purpose: Compared amount, type, order of information to other sources; Setting: 4 units in 2 hospitals, UK. Sample: Convenience; 5 handoffs, 3 patients, random sample 60 charts. Handoff method: verbal, office; Care model: Team nursing | Design: Descriptive, comparative, 2 × 2 design. Methods: Content analysis. Multidimensional scalogram analysis. Reliability, validity assessed | More information in notes than handoff. Global judgments in handoffs only. First 4–5 items ordered the same; varied after that. Information varies by specialty |
|
Nagpal et al. (2010b) Study #2 |
Purpose: Identify problems, needed components. Setting: GI department, teaching hospital, UK. Sample: Convenience, 18 physicians, nurses. 50 nurses, physicians in Delphi Handoff method: Unknown; Care model: Not described | Design: Qualitative, descriptive (Delphi). Methods: Interviews, Delphi technique | Consensus list of postoperative handoff data elements Categories are patient information, anaesthetic information and surgical information |
| Staggers and Jennings (2009) | Purpose: Handoff content, context. Setting: 7 medical, surgical units, 3 hospitals, US. Sample: Purposive, 13 sessions, 38 nurses, 53 patients. Handoff method: verbal, taped, bedside; Care model: Not described | Design: Qualitative, focused ethnography. Methods: Handoffs audio taped, transcribed, field notes. Content analysis | 4 themes: The Dance of Report, Just the Facts, Professional Nursing Practice, Lightening the Load. No structure, high noise levels, interruptions, no EHR use |
| Welsh et al. (2010) | Purpose: Barriers, facilitators. Setting: 3 units Veteran's Administration, US. Sample: Convenience, 20 nurses; Handoff method: taped, written; Care model: Not discussed | Design: Qualitative, descriptive. Methods: Interviews. Grounded theory | Barriers – interruptions, info amount, quality, questions, equipment; facilitators-pertinent info, interaction, structured |
Results
The final set of relevant studies totals 30, displayed in an evidence table (Table 1) and as detailed, supplementary web material (See supporting information Table S3 in the online version of the article in Wiley Online Library).
Overview of studies
The set includes 20 qualitative, 4 experimental and 6 descriptive studies. Ethnographic approaches were employed in 10 of the qualitative studies with a sampling of other techniques – unspecified with content analysis (4), grounded theory (2), case study (2) phenomenology (1) and discourse analysis (1). The most frequent experimental study design was pre-experimental (3); the other study employed 2 × 2 factorial design.
The Bowling quality appraisal and QUASII ratings are mirrored evaluations for experimental studies even though they addressed different aspects of the studies. Overall, none discussed statistical power and only one included sufficient material about the sample and instruments. All discussed study implications. The QUASII scores ranged from 11–29 with 1 scoring 11; 2 in the middle range of scores and only 1 had a score above 21. The Bowling quality appraisal shows higher ratings overall for descriptive and qualitative studies. Common issues included not listing study limitations or discussing statistical power where applicable.
Authors conducted studies most frequently in the UK (14), then the US (7), Australia (5), Scandinavia (3) and a single study was completed in Israel. All used convenience samples of nurses except Lamond (2000).
The majority of authors did not list a definition for the handoff event. The UK and Australian authors used the term ‘handover’ while US authors used ‘handoff’. For those defining the term, no consensus among definitions is evident across the 30 studies.
The majority of authors (20) did not specify an underlying theory for their studies. Of those who did, authors used information processing/schema theory (2), change management theory (1), symbolic interaction theory (2), psychoanalytic theory on rituals (1), action research (1), linear communication (1), ‘information overload’ (1) and knowledge concepts from Gadamer and Ricoeur (1). All researchers examined handoffs from the perspective of nurses except for two authors (Berkenstadt et al. 2008, Benham-Hutchins & Effken 2010) who traced patient-centred handoffs across hospitalizations.
Information effectiveness and efficiency
One third of the studies (10) addressed information effectiveness and efficiency, i.e. information completeness, accuracy, efficiency (time for handoffs and costs). Authors of descriptive studies found incomplete information in verbal handoffs compared to formal documentation (Sexton et al. 2004) or taped report (O'Connell & Penney 2001). Authors noted information loss at every handoff point in the care continuum for surgical patients (Nagpal et al. 2010b) until only 43·9% of essential information reached the inpatient unit. Likewise, Richard (1988) found incongruence in all 57 handoffs for 584 patients when comparing handoff information to patients' actual conditions.
Structured, consistent formats improved information completeness using different formats (Pothier et al. 2005, Nelson & Massey 2010). Bedside handoffs were difficult on units such as palliative care because of the sensitive nature of information being conveyed (Mckenna & Walsh 1997). Taped handoffs had more information omissions than verbal reports but were more congruent with the patient's actual condition according to a study completed over 20 years ago (Richard 1988).
Pothier et al. (2005) discovered that nurses in a laboratory experiment had only 31% accuracy on traditional written notes after handing off information five times. A combination of verbal report with a typed sheet resulted in minimal information decrements. Dowding (2001) found low average information recall rates of 27% and only a 45·2% average accuracy rate during a laboratory study comparing two handoffs with different information structures. However, she also found that information recall and accuracy were improved with consistent formats.
Authors showed that average times and costs for handoffs decreased with a standardized, electronic template (Nelson & Massey 2010) and methods tailored to each unit (Mckenna & Walsh 1997) except with a format called ‘Essence of Care’ (Fenton 2006).
Handoff processes and structures
Authors of nine studies examined the structure and processes of nursing handoffs. The majority outlined the functions of handoffs: transfer of patient information, team building, nursing goals/values, social, education (Lally 1999), consensus about next shift care actions and knowledge construction (James et al. 2010), group cohesion (Payne et al. 2000), expectations and cultural aspects (Holland 1993), emotional support (Hopkinson 2002) and socialization of new nurses (Ekman & Segesten 1995). Handoffs are deemed complex activities (Sherlock 1995, Benham-Hutchins & Effken 2010) and handoffs can be cryptic, given at high speed and can include jargon (Wolf 1989, Payne et al. 2000).
Other authors addressed unique topics such as patient-centred, multi-professional handoffs using social network analysis (Benham-Hutchins & Effken 2010). Networks of providers ranged from 11–20, were nonlinear, unpredictable and shaped by providers' information needs. Other authors concluded that replacing face-to-face handoffs with asynchronous electronic records would hinder nursing handoffs (Engesmo & Tjora 2006). Chaboyer et al. (2010) discovered that patients participated in less than one half of bedside handoffs and that nurses' printed sheets played a primary role during and after handoffs.
Handoff methods varied: verbal, office-based (10); several described methods (8); taped (1); reading documentation (1), bedside (1), but many authors did not mention handoff methods (9). Units employed team nursing (5), primary nursing (3), a combination (2), ‘pairs’ (1), but the majority did not discuss care models (16). Three were laboratory studies (Dowding 2001, Pothier et al. 2005, Berkenstadt et al. 2008) and a unit care model was not applicable.
Handoffs as ritual
Authors of six studies examined handoffs as rituals. The first work was by Wolf (1988, 1989) but work has continued as late as 2008 (Evans et al. 2008). Consensus exists across all authors that handoffs meet the criteria to be considered a form of ritual. Rituals have a set of common values and symbolic language ensuring exclusivity. They can alleviate anxiety (Evans et al. 2008), serve valuable psychological and social protective functions (Strange 1996) and provide a safe forum for nurses (Wolf 1988, 1989). Strange concluded that handoffs had important ritualistic functions that should not be dismissed.
Information content
Only five studies addressed the information content of handoffs; however, purposes and findings are dissimilar. Hardey et al. (2000) outlined the crucial role that nurses' papers or ‘scraps’ played during and after report for information communication and care delivery. Lamond (2000) examined the information content of handoffs and other sources, concluding that although more information exists in documentation, nurses communicated global judgments only in handoffs. Nagpal et al.(2010a) published described general information to support patient transfers across a surgical product line. Staggers and Jennings (2009) found four themes during handoffs with facts and professional judgments being only second and third in frequency, handoffs were unstructured, replete with interruptions and high noise levels and nurses did not use available electronic health records. Last, Welsh et al. (2010) identified barriers and facilitators of written and taped handoffs.
Discussion
The concept of handoff
The majority of researchers of medical and surgical handoffs did not specifically define the term used for handoffs and of those who did, no consistent definition is apparent. This finding is congruent with Cohen and Hilligoss (2010). However, definitions are emerging (Solet et al. 2005, Cohen & Hilligoss 2010). The event could benefit from greater concept clarification, consistent with work completed on handoff functions.
Functions
Authors agree that handoffs are complex (Parker et al. 1992, Sherlock 1995, Strange 1996, Benham-Hutchins & Effken 2010). Researchers outlined the various functions of handoffs, concluding that handoffs perform crucial functions for nurses. As was noted, this event performs no small suite of functions extending far beyond information transfer including educational, role socialization, emotional, social and other functions. These aspects of handoffs are now well outlined having been the topic of numerous studies. What is clear is that handoffs, especially face-to-face handoffs, contribute to nursing and should be retained because of these implicit purposes. Even as handoffs become computerized, handoffs should not be eliminated, being instead supported and not replaced by technology. Moreover, handoffs can serve as a form of resilience, as mentioned earlier (Patterson & Wears 2010) to detect erroneous assumptions, actions and errors.
Information
Researchers attended less to the informational aspects of handoffs despite information being its purported cornerstone. Researchers described informational issues with current processes in the degradation of the amount of key information across handoffs (Pothier et al. 2005, Nagpal et al. 2010a) and incongruence between information and patients' conditions (Richard 1988). Only a minority of study authors (6 of 30) addressed information content for medical and surgical settings.
Solutions to incomplete and inaccurate information are emerging. Researchers found that structured, consistent formats improved information completeness although the formats were all different and, as Lamond (2000) noted earlier, typically atheoretical.
Handoff structures are a current topic of global interest. Australian clinicians are conducting a national project to standardize handoffs (Mcmurray et al. 2010). Nagpal et al. (2010b) developed a tool for inter-unit transfers for gastroenterology patients. In the USA the format S-BAR (Situation, Background, Assessment and Recommendation) is being touted as the structure of choice. However, this structure is controversial because it was developed as a communication method between professions (nurses and physicians) primarily to report patient care issues; its format must be highly tailored to fit nursing handoffs. No studies as yet addressed S-BAR efficiency or effectiveness for medical and surgical units. Earlier reviews also indicated that research is absent on its S-BAR effectiveness generally (Cohen & Hilligoss 2010, Riesenberg et al. 2010). More important, Winters and Dorman (2006) stated that any national efforts to improve patient safety should be supported by sufficiently strong evidence to warrant such a commitment of resources. This should be the case with S-BAR or any other format being considered.
What may be emerging is that the specific structure may be less important than having a structure for a contextually based handoff. Wilson et al. (2007) suggests each unit or subgroup should agree on a specific format. This is in contrast to the effort in Australia to adopt a bedside handoff method and standardize content (Mcmurray et al. 2010) and to the S-BAR adoption. What may be most important is the consistency and predictability of information transfer for day-to-day operations on particular patient care units. This notion would need to be tested; however, the concept is congruent with findings from work in psychology and human-computer interaction (Nielsen 1993, Zhang et al. 2003, Shneiderman et al. 2010).
The content of report has been researched since at least 1969 (Clair & Trussell 1969), but clear guidance on information for medical and surgical units is not apparent from available research. Nagpal et al. (2010b) offer general, interdisciplinary information for transferring surgical patients from unit to unit, but more detail is needed to support intra-unit, nursing-specific handoffs. Global judgments about patients are communicated in handoffs but not in documentation (Lamond 2000) and the majority of handoff content in another study was related to communication itself vs. clinical content (Staggers & Jennings 2009). Last, Hardey et al. (2000) documented the crucial role of nurses' papers called ‘scraps’. Work is needed to establish information content for handoffs, especially key information for nurses receiving handoffs for patients they have cared for previously vs. patients that are new to them, and whether or not nurses will be amenable to technology-supported handoffs given their views on the primacy of paper forms (Staggers et al. 2012).
Handoff methods
No one handoff method emerged as more effective and efficient. Bedside handoffs, although popular, can be a difficult approach on some units because of the sensitive information being transferred (e.g. palliative care, psychiatry or oncology) or due to patient conditions (those prone to infections such as those in transplant units). A more promising method was a verbal handoff supplemented with a printed form, but this method was tested only in one nursing study. (O'connell & Penney 2001) concluded that no one type of handoff was more effective because of differences in contexts, patients and nurses.
Handoff context
Most studies did not address the impact of existing care models; while this might impact generalizability, the authors' current research (Staggers et al. 2011, 2012) indicates that the defining factors include tailoring for patient-centred and nurse-centred aspects. Handoff context has not been sufficiently considered, until very recently (Mayor et al. 2011, Staggers et al. 2012) despite its importance. Mayor et al. found that task uncertainty impacted handoff duration. Moreover, making conclusions across settings ignores differences in work design, patient conditions and nurses. At minimum, future researchers need to organize reviews by unit type.
Considering context is critical to successful computerization (Staggers et al. 2012). Contextual factors pertinent to handoffs include at least patient condition, nurse expertise and handoff method. For example, patient units may combine medical and surgical patients, but their handoff content is quite distinct. Thus, forcing specific content and handoff methods across all contexts is not defensible. Also, unit-specific and cross-unit content needs by patient type need to be defined. Precise content is needed to support nurses' decision-making, planning for care, and for detecting critical information.
Handoffs as rituals
Researchers completed a set of studies about handoffs as a ritual. Although it may be tempting to dismiss handoffs as mere antiquated rituals, these authors give insight into the positive role of rituals. That is, handoffs are a cultural phenomenon with each unit having norms, expectations and codes of behaviour. Handoffs actually play an important role in nursing by serving psychological and social protective functions.
Research and methodological issues
The majority of authors completed qualitative studies on medical and surgical units. Although others recommend increased quantitative work (Riesenberg et al. 2010), we applaud the amount of qualitative work completed to date and encourage researchers to continue use of this method to expand understanding of the concept of handoffs.
The overall quality of the quantitative studies to date for medical and surgical units is low. The low quality of quantitative studies is consistent with a previous review of 20 cross-context studies from the USA (Riesenberg et al. 2010). This previous analysis and the one here have only one study in common to both reviews. Thus, the previous findings are extended now to international studies that exclude nurses' perceptions about handoffs. Four major issues contributed to this finding in the current review. Most study designs were pre-experimental with the exception of work by Dowding. These designs did not allow for control of potential confounders, limiting the validity of study findings and their generalizability. Second, researchers did not account for potential differences among nurses, their characteristics or setting including knowledge, expertise or other attributes, unit leadership that might affect study findings or differences in settings. Researchers typically used convenience samples of nurses. Third, potential confounders were not included in statistical analyses. Last, studies were not generalizable due to contexts and sample sizes. Studies were cross-sectional in design. None were longitudinal in a naturalistic setting.
Unit of analysis issues
All but two studies examined nurses as the unit of analysis for handoffs. Two sets of authors traced handoffs from patients' perspective across a hospital stay (Benham-Hutchins & Effken 2010, Nagpal et al. 2010a). Both perspectives are needed; however, researchers will want to carefully consider which one in future research. For instance, nurse-centred handoff guidelines are emerging. General guidelines do not consider how well the nurses ‘know’ the patient or nurses’ expertise.
A critical point, especially for quantitative study designs, is made by Karsh and Brown (2010). Researchers need to consider how study variables at different levels of the organization, unit, and provider might affect errors, patient safety (or handoffs). For these, nested designs are most appropriate, but not considered yet in handoff research.
Future research
Two areas of study are saturated: handoff implicit functions and rituals. Handoff methods have not yielded clear results despite the number of completed projects and studies. Comparative studies are rare except comparing traditional report to any new method. A focus on the informational aspects and content of handoffs is greatly needed, a finding consistent with Matic et al. (2011). How can researchers enhance the positive aspects of handoffs like resilience (Patterson & Wears 2010)? How can computerized designs augment nurses' cognitive to support resilience? Does a more accurate and complete handoff result in better patient outcomes?
More research is needed on patient-centred handoffs. How can handoffs from the patient's perspective be improved to prevent information degradation across units and settings? Also, computerized designs are emerging; however, these are also untested as yet.
Future research related to context is imperative. Understanding information content related to patients' conditions, nurses' expertise and handoff content based on contextual variables could improve future handoffs. Researchers also need to be cautious to report existing care models, handoff methods and other contextual attributes important for generalizing findings.
Improvements are needed on research design for quantitative studies, considerations to differences in nurses, contexts and statistical analyses. Pre-post test and post test only designs are weak; future researchers should consider other designs if at all possible. Other designs are available (Shadish et al. 2002) for use in actual settings such as quasi-experimental designs, matching participants' pre- and post interventions and longitudinal and time-series approaches. Experimental studies can, and have, been tested in simulations, an approach that protects patient safety. Considering handoffs from patients' perspectives over time is a promising approach. Past studies typically used providers' perspectives in a cross-sectional design. More sophisticated sampling techniques can be used rather than convenience samples. These techniques could include random samples of providers or patients or stratified (by experience or education) samples.
Limitations
This study was limited to publications in English and in peer-reviewed journals. The quality of quantitative studies was not high and few experimental studies were available.
Conclusion
The implications of this work are several including: (1) research funding, (2) nurse executives and (3) clinical nurses. First, it is critical that funding be made available to support nursing research on handoffs. The emphasis to date has better supported research on physician handoffs, particularly with the reduction in resident working hours in the USA and UK. Second, nurse executives and managers can be made aware that the current popularity of bedside handoffs is not yet supported by quantitative research on medical and surgical units, some settings are not amendable to this handoff method and handoffs serve multiple functions beyond information transfer, including error detection. Last, clinical nurses can be active in the transformation of handoffs tailored to their own contexts and in determining information critical to accurate and complete handoffs.
Acknowledgements
The authors would like to thank Dr Marge Benham-Hutchins for her thoughtful review of a previous draft of this manuscript. We would also like to thank research assistants Denise Chesney and Emily Mlupi for their work on the quality appraisals.
Funding
This work was supported by an internal research grant from the College of Nursing, University of Utah.
Conflict of interest
The authors have no conflicts of interest.
Author contributions
- substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data;
- drafting the article or revising it critically for important intellectual content.
