A tool to assess the components of a safe and quality medical handover

Study design, setting and data collection

This observational, prospective study occurred at a 600-bed tertiary hospital in Australia. The 4 pm after-hours team handover was selected for this study, because it occurs daily at a designated time and location.

An initial literature search was performed using keywords including, “medical handover”, “handover”, and “patient safety” through the PubMed database to identify published articles relating to medical handovers. In addition, a wider internet search was also incorporated to include local and international guidelines and policy articles focused on best practice in clinical handovers.^[1,3,7,8]

Based on common themes in the literature, we extracted the vital components of a medical handover. These components were subsequently refined to suit the after-hours handover in our hospital, and incorporated into our tool. These ultimately included team members involved, environment and logistics of handover, structure and content, management plans, the patient, documentation and teaching opportunities. The questionnaire was trialled at one clinical handover, and no changes were made to the tool prior to the study period (Table 1).

Data collectors attended 20 after-hours team handovers, over a two-month period in 2022. Data collectors were medical professionals and were not involved in the handover themselves. Random handovers between Monday to Friday were attended, dependent on data collector availability. There were no unique or confounding factors, such as COVID, which influenced the usual practice of this handover. At the handover, the data collection tool was completed and submitted electronically to a REDCap database.^[9,10] This data was subsequently analyzed as described below, and strengths and limitations of the tool were elicited in the analysis phase of the audit by the authors.

Statistical analysis

Data was exported from REDCap to Excel, and subsequent analysis was performed by the study team.

Members involved in the handover

Of the 20 after-hours team handovers, 65% had the expected number of attendees—30% of handovers were short on Resident Medical Officers (RMOs), and 50% of handovers were short on nursing staff. The handover expected one Registrar, four RMOs and three nursing specialists per handover. We found an average of 1.25 Registrars, 3.7 RMOs and 2.5 nursing specialists attended each handover. Non-attendance was due to staff sickness and alternative clinical commitments. There was an obvious leader in all handovers, and it was always the Registrar. Jobs and responsibilities were always discussed and allocated to members of the team during the handover.

Environment and logistics

All handovers occurred in a fixed location, and in a quiet environment. 85% of handovers experienced an interruption—60% interrupted by phone call(s), 45% interrupted by pager(s), 15% interrupted by other (including MET calls, technology issues) and zero interrupted by people.

Structure and content

No patient handover strictly followed an accepted handover structure e.g., iSoBAR. Given the handover is based off the electronic POC list as previously described, all patients were prioritised. In 20% of handovers, there were patients where it was not entirely known why they were a POC. For example, a patient was listed as a POC for a recent MET call, but the reason for the MET call was not described, for example tachycardia or hypoxia.

Management plans

Due to lacking or brief electronic patient data, there was never access to completely up-to-date summaries of patient admissions, nor access to clear, up-to-date management plans for each patient.

Patients

Patient privacy was always ensured, i.e., not within hearing distance of other patients and visitors. No patient was involved in any of the handovers.

Documentation

There was no documentation of handover itself having occurred in the patient clinical records.

Teaching

Teaching opportunities were found in 10% of the handovers.

Statement of principal findings

Through the tool, it was found that the handover environment and patient privacy were optimal. Moreover, several areas for improvement were identified; including staffing shortages, repeated interruptions, lack of a structured handover and poor access to up-to-date management plans. The tool itself successfully allowed analysis of key medical handover components; however, it did have some limitations.

Strengths and limitations

Strengths of this study included the development of a tool which efficaciously gained insight into medical handovers, particularly at our hospital where medical handovers had not been previously evaluated. Bias was limited with data collection being carried out by members not involved in the handover process itself. However, generalisability is limited by the small sample size of 20 handovers. Additionally, the nature of this medical handover was unique, with a team eliciting information from computer software, thus limiting generalisability to person-to-person handover. Consequently, areas for improvement include widening the audit scope to assess other team or ward handovers, at different times of the day or week.

Limitations in the tool itself became apparent during the analysis phase. Chiefly, the questionnaire did not capture the number of patients discussed at each handover. While initially this did not seem to be a key component to include within the tool, its lack narrowed the clinical correlation of some findings. For example, the audit was able to show that 20% of handovers included patients whose reason for being a POC was not clearly identified. However, given there was no data captured on the number of patients discussed at each handover, it prevented this from being expressed as a proportion of patients in which this was a concern, thus potentially overestimating the number of patient handovers affected.

The tool was created with tick box and drop-down menu options for most questions. While this enabled ease of data collection, it limited the capacity to clearly reflect the extent of certain issues identified. For example, the audit was able to show that 85% of handovers experienced an interruption; however, unless specifically commented upon under additional comments, this did not allow for differentiation between a handover that was interrupted multiple times, and a handover that experienced a more minor interruption. We attempted to reduce this limitation by having free text boxes throughout the questionnaire; however, in retrospect there should have been another question explicitly regarding the number of times the handover was interrupted.

Interpretation within the context of the wider literature

Our tool performed well by allowing analysis of the key components of medical handover and highlighting the components which were optimal, and those which require improvement.

In terms of handover attendance, it was found that 65% of handovers had the expected number of attendees. It is known that staff shortages and high workloads compromise patient safety,^[11] and thus hospital administration must implement mechanisms to prevent or manage staff shortages. Ideally, 100% attendance should be achieved, as staff are rostered on the shift and expected to attend the handover. An obvious leader was always present, and there was always an agreed allocation of jobs and responsibilities. Importantly, the Registrar was the leader in every handover—senior clinician involvement is essential as it ensures the appropriate level of management decisions.^[10] Whilst teaching is not a fundamental component of handovers, 10% of handovers contained teaching. This included discussion of interesting conditions presented and management strategies.

In regards to patient safety, patient details and prioritisation were satisfactory for safe handover, however; improvements in documentation could be made. Patients’ details were always sufficient for safe handover; including patient hospital number, full name and date of birth. This met minimum standards set by guidelines, and is an integral part of any handover.^[3,7,8,12] There was always prioritisation of patients handed over, since the handover solely includes POC. Beyond this, there was no specific order to the patients discussed; however, unwell patients were sometimes further prioritised, to allow the Registrar to review them immediately after handover. Prioritisation is important to ensure unstable patients are reviewed early, and by a senior doctor.^[13] Regarding improvements in patient safety, there was no documentation in the patient records of the handover itself having occurred. However, when a patient was reviewed following the handover, the doctor would document their assessment which is essential. It is not common practice to document handover of care officially; rather it is assumed that the team receiving handover will adopt responsibility.

The environment was suitable for safe handover, being quiet, private and in a fixed location, however; improvements should be made to prevent interruptions, given 85% of handovers were interrupted. In our hospital, all emergencies (medical, personal, fire, etc.) are transmitted through every pager, affecting the quality of the handover. It is recommended that non-critical interruptions are limited.^[3] Therefore, the authors recommended that handover be “pager-free”, excepting life-threatening emergencies,^[7] or allocating one person to receive all pages. Alternatively, we recommend that personal phones are allocated to staff instead of pagers, to provide a more direct method of communication, thus limiting unnecessary interruptions.

The iSoBAR framework, endorsed by the World Health Organisation,^[5] provides a standardised approach to handover structure. This audit showed that none of the handovers strictly followed an accepted handover structure. It is essential that good handover practices are revisited and maintained; however, education regarding handover may not be prioritised in hospitals due to time constraints.^[14] Protected teaching time, and accessible resources are effective ways to counteract this.^[14] Although a consistent and clear structure is recommended, it is suggested that the “iSoBAR tool should be used in a manner that suits the clinical context to guide the content and structure of the handover”.^[3] Ultimately, the handover studied is unique given its electronic format, and thus a distinct iSoBAR handover may not be reasonable when the medical staff are yet to review the patient.

No handovers had access to wholly up-to-date summaries or management plans for every patient. As aforementioned, our after-hours handover is unique in that there is not one medical team handing over to another team, rather the team elicits information from the hospital-wide software POC list. Accurate and safe handover depends on the handover software being kept up-to-date with clinical progress, and this did not always occur diligently. This is a major shortcoming, and thus we submitted several recommendations to our health care service. We recommended educational resources to be placed on the wards to notify medical staff of this downfall. Additionally, the task of updating the electronic medical handover must be allocated to a staff member, perhaps a RMO, to ensure the electronic record is updated prior to handover. Furthermore, we recommended changes to the electronic software to ensure that minimum information in each domain of iSoBAR must be entered before the handover can be submitted.

Finally, 20% of handovers experienced a failure to identify why a patient was a POC. When a patient is added to the POC list, staff must input a reason for concern—be that MET Call or ICU step-down. In 20% of handovers, there were patients who appeared on the list, who either—did not have MET Call or ICU listed, or if they did, there was no elaboration as to why, or what had happened. For example, they may have listed MET call but did not elaborate if the emergency was due to hypoxia, hypotension or arrhythmia. To ensure more information was available for handover, we similarly recommended an alteration to the electronic software to guarantee that as “MET Call” is input as the reason for POC, a clinical scenario such as hypoxia must also be entered. In medical handovers, there is an expectation that “minimum information content” is reached.^[1] This includes diagnosis, clinical condition, up-to-date management plans, and emerging critical information.^[1,12] Our handover system includes free text boxes, with no minimum information required. Studies have demonstrated that free-text boxes are often insufficient for safe handover, and to be seen as useful by staff, electronic handovers must be simple, practical and informative.^[12] Without direction, essential information for a safe handover is frequently omitted, which has implications for patient safety.^[15] If a health service relies on electronic handovers, we recommend designing software that requires a certain number of details to be included in order to be submitted, thereby aiding handover safety and effectiveness.

Implications for policy, practice and research

This audit highlights practical improvements for our hospital, as well as broader implications for other healthcare services. Staffing shortages are known to impact patient safety,^[8,11] and we recommend ensuring adequate on-call staffing to cover sick calls. Minimizing repeated interruptions to handover would improve the efficiency and safety of handovers. Implementing this could appear different in different health services, for example, designating “pager-free” time,^[8] limiting non-critical interruptions,^[3] or allocating one person to receive all messages. Finally, sufficient details of patient admission are required for safe and effective handovers. There are benefits to an electronic medical handover, such as legibility, providing more in-depth information than written handovers, and improved staff satisfaction.^[16,17] Despite these benefits, our data demonstrates difficulties in accessing up-to-date management plans and patients’ current clinical status. As discussed, mechanisms must be implemented in a healthcare service to ensure minimum content information is reached, in both verbal and electronic handovers. Ultimately, if a health service has not reviewed its medical handovers, our study raises the possibility that there are areas for improvement within their practice.

In this regard, our audit questionnaire is a useful tool for health services aiming to audit the medical handovers within their organisation. The questionnaire was able to successfully capture a breadth of information related to a number of key components of medical handovers. However, through analysis of the data, it was also clear that the audit tool was limited in certain aspects, such as the lack of data reflecting the number of patients discussed at each handover, and the restrictive nature of the tick box format as aforementioned. As such, the questionnaire developed and tested by our audit could be utilized by other health services as a starting point for the development of their own audit tool, by adapting the strengths of our questionnaire and building upon the areas for improvement. Finally, a further project with wider scope could assess medical handovers at various points of the patient care journey, including transfer of service or discharge to other facilities.

In conclusion, the optimal handover may not be a one size fits all approach; however, it is important to be aware of minimum standards for a medical handover, and ensure these are implemented within the context of the healthcare service. Through the development and application of our audit tool, we have identified areas for improvement in our after-hours medical handover. Looking forward, this tool can be adapted by other healthcare services aiming to evaluate the medical handovers within their organisations. This tool can be easily adapted by utilizing the headings and questions, as well as some of our recommendations, to generate an audit tool suited to different handover environments. As with our healthcare service, we expect that if medical handovers were evaluated in more international hospitals, improvements in medical handovers would be elicited, and ultimately patient care and safety could be further optimized.