9.3Identify

9.3.1Patient identification

Identification of the patient is a fundamental and frequently undertaken task that sets the foundation for ongoing care. While the routine nature of this act can diminish its perceived relevance compared to more cognitively demanding clinical tasks, the potential role and impact of incorrect identification in the escalation of clinical errors should not be ignored (ACSQHC 2012a). Approved identifiers include patient name (family and given), date of birth, gender, address, medical record number and individual health care identifier. The Australian national standard on patient identification stipulates that ‘a patient’s identity is confirmed using three approved patient identifiers when transferring responsibility for care’ (2012a, p. 17). It was not possible to determine how many personal identifiers were specified in the patient expect call, or at which point complete patient identification and demographic information was entered into the patient electronic records. However, it is realistic to expect that only the minimum required details would be communicated in the original patient expect call and that more comprehensive details would be entered during admission and registration. The patient’s name, age, date of birth, gender, address and medical record number were all available as headings in the electronic record and were generated on both the Triage Nursing Assessment form and hospital identification stickers.

Examples of typical patient identification errors can be found in Shirley’s records. For example, only two identifiers were often used at the beginning of handwritten notes: age and gender. During the patient transfer, identification codes and case numbers were specific to the service provider and did not represent a point of continuity. The use of patient identification stickers with the patient’s address and birth date provided the approved descriptors but this information was not always consistent with hand notations made on the documents. This incongruence indicates that the stickers were either not in place at the time the notations were made or not attended to by the scribe.

To further illustrate mistakes, in the documents for ambulance, Royal Flying Doctor Service, and medical notations in the emergency department admission, Shirley was identified as a 47-year-old female, 85 [female], 74-year-old female and a 74 [male]. These inaccuracies were due to the incorrect use of symbols, age calculation errors and digit transposition. While seemingly innocuous, they show the potential for errors of commission in even the simplest of communications involving symbols and digits. The typical use of the patient’s age as a personal descriptor in notes taken during interviews is both logical and expedient in terms of contextualizing or framing the patient’s status but does not comply with standards for identification. Furthermore, age-related identification errors, if not picked up, may inadvertently lead to different treatment protocols being applied or admissions to incorrect wards.

9.3.2Clinician identification

Two essential terms used in the definition of clinical handover are responsibility and accountability. For example, Jeffcott et al.’s (2009) framework for defining and measuring handover mentions three key handover elements: information, responsibility and/ or accountability and system. However, Cornock (2011) argued that while the terms responsibility and accountability are often used interchangeably, they have distinct legal meanings and ramifications. A second aspect of identification investigated in the documentation audit therefore concerned determining responsibility and accountability for patient care. Without belaboring the legalistic distinctions, being able to easily identify the names and designations of the health workers providing care at any specific time is fundamental to understanding who is personally responsible for tasks undertaken and who may be held accountable for the decisions made and directions specified for a patient’s care. In this instance Australian Medical Association and hospital standards for medical and nursing entries in medical documentation were used as the audit reference. Given the situational and temporal demands of a rural transfer, being able to follow the chain of responsibility and accountability for care across a rural patient’s interhospital journey can be problematic. We will now discuss three points that emerged from our review of the electronic record and documentation audit: determining clinical responsibility and accountability, diffusion of personal responsibility and delegation of clinical responsibility.

9.3.3Determining clinical responsibility and accountability

Consistent with the traditional medical model of care, the sequencing diagram in Fig. 9.4 illustrates how the handover of clinical responsibility and accountability from a medico-legal perspective can be separated from handover of patient care in the more generic sense. For example, in Royal Flying Doctor transfer standards, handover of responsibility and accountability remains doctor to doctor. If a flight doctor is not in attendance during a patient transfer only partial aspects of care are handed over to the flight nurse. In this instance ultimate responsibility for the patient and accountability for actions taken remains with the referring clinician until such time as the patient is handed over to the receiving clinical team at the emergency department.

The specific clinician’s level of expertise and scope of practice will determine their capacity to make decisions and act on either their own authority or through delegated authority. The timing of a patient enquiry or a patient expect call can therefore impact on a patient’s treatment episode, as who is available to discuss and determine the patient’s treatment and admission status is often timing dependent. If shift changes occur during the transition of care, the clinicians who initiated the process may not be available at the conclusion.

In our audit of the electronic records we found it difficult to determine who made either the initial patient inquiry or expect call, who received either of the calls or who was accepting responsibility for the patient at the receiving site. When creating records, standards require that clinicians provide legible identification details including date and time of interaction, the doctor’s name or a legible signature, clinical level, speciality and contact details of the clinician. Similar requirements are expected for nursing and allied health notations. In the case of the patient expect call, identification details of the receiving personnel were noted in only 4.7% of the patient expect records (twice by first name and four instances by initials). The time the patient expect call was received populates a separate heading in the electronic record. Notations in this heading indicated that 55 calls (43%) were received outside normal working hours when access to senior medical officers and consultants was limited.

Fig. 9.4: Chain of clinical responsibility

The name of a clinician at the receiving hospital was provided in 14 records (11.1%). Both the accepting doctor’s name and designation were provided in only three of these cases (e.g. GEN SURG ACCEPT REG NAME). In a further two cases the name of the consultant under whom the patient was to be admitted was clearly specified by the phrase ADMIT UNDER DR NAME and GASTRO CONSULTANT DR NAME ACCEPTED PT. A pager number to contact a medical officer in the specialist area on the patient’s arrival was provided in four cases (3.1%) and comments such as MED REG AWARE, ENT SURGEONS AWARE or KNOW TO DR NAME were noted in a further six cases.

The different terminology used in these notations reflects the diffusion and delegation of authority that occurred. The capacity to stipulate the admitting consultant is dependent on an array of factors. As indicated above, one of these factors is the timing of both the original clinical inquiry and the patient expect call. The day (weekday or weekend), time of the calls (in or after hours), and staff rostering have implications for the availability of individual consultants and specialist admitting teams at, and subsequent to, the patient’s arrival. For example, a registrar can accept a patient on behalf of a specialist medical unit but an admission must be under a specific consultant. However, the accepting registrar may be aware that they will not be on duty when the patient is scheduled to arrive, rendering redundant their nomination as the person to contact on arrival. As the names and designations of the medical officers involved were rarely provided for either the sending or receiving sites, it was difficult to determine who held responsibility for the patient during the transfer process.

A lack of compliance with identification protocols was also evident in the medical records. In many patient records the capacity to identify who did what and when was problematic, including in the emergency department Triage Nursing Assessment document in which multiple entries are made by different nurses over the duration of the patient’s emergency department stay. Despite clear instructions on the triage document for contributors to provide a signature, initials, printed surname and title (designation), inadequate space was provided for clinicians to do so, resulting in low levels of compliance and poor legibility. Nurses’ initials were often not traceable back to their signatures and a legible printed name was not identified in any of the audited documents of this type.

Similarly, medical notes by both emergency department and specialist clinicians in the emergency department Integrated Admission documents rarely complied with all aspects of the required reporting procedures. Omissions included legible signatures and/ or printed names, the time and date of the entry, designators and most frequently contact details. The identification of clinicians in both ambulance and Royal Flying Doctor Service records showed a high level of compliance, expedited by the use of ID codes in the ambulance service and by the provision of adequate space for clinician details in the Royal Flying Doctor Service documents. The absence of complete clinician identification details appears to be linked to design faults in documents, time constraints and misplaced assumptions that clinicians would be identifiable. The lack of personal detail may also be linked to the accepted practice of diffusing responsibility from the individual to a medical team, unit or agency as discussed below.

9.3.4Diffusion of personal responsibility and accountability

The patient expect notations indicated that the responsibility for care was generally diffused to a site or treating area rather than to a specific clinician. This reflects the prevalent and pragmatic reality of the team-based care models that operate in public hospitals. The initiating clinician was identified by name in only one patient expect record, while the initiating site was noted in 72 cases (56.7%). The accepting clinical unit was specifically noted for 99 cases (77.9%), while an accepting physician from the initial clinical inquiry was specified by name in only 14 cases (11.1%). This trend continued across the medical records in our audit. The more complex the case, the more specialties and clinicians involved in the patient treatment, the more limited the likelihood appeared to be of a lead clinician being identified in initial phases of treatment.

A further issue was the frequent use of nonspecific identifiers in discharge documentation. These identifiers include terms such as DR D ATTENDING, DR G NIL, DR T MEDICAL OFFICER or DR G UNKNOWN. Such descriptors were used in 50% of the discharge letters audited. When a discharge letter was directed back to a rural health facility or general practitioner group and the clinician’s name was not known, the system prerequisite for compulsory fields to be populated in the generation of electronic documents required that sending clinicians adopt a work-around strategy, or in human factors terms, conduct a routine violation. Given the age and nature of the diagnoses of many of the patients included in this small sample audit, it was surprising to find that 9 of the 20 patients did not have or could not identify a ‘Nominated Healthcare Provider’ for their post-discharge care. Difficulties associated with health care workforce management in country areas may be contributing to this situation.

9.3.5Delegation of responsibility and accountability

The issue of delegating responsibility was seen in the patient expect call when junior medical officers and registrars accepted patient transfers on behalf of consultants in specialist areas. However, as indicated above, this was rarely made explicit. The relevance of delegation became more apparent in the stage 3 audit. The common and well-accepted practice is for junior medical officers to complete handover tasks and documentation on behalf of more senior medical officers and consultants. In particular, junior medical officers generally write the ward round notes and discharge documents. However, in the documents we audited it was difficult to determine the level of supervision of this delegation of authority. We would argue that the handover of the patient’s care and responsibility should be considered with regard to the distribution and use of power hierarchies in the clinical setting and that the explicit identification of individual clinicians in documentation at transition points of care across the patient journey warrants greater consideration.

To complete this section we will briefly summarize how identification issues manifested in Shirley’s story. As a starting point, neither the electronic notations nor Shirley’s written records indicated the names of the clinicians who made the initial patient enquiry call to the orthopedic unit, those who accepted the referral or those who made and received the patient expect call. As a result, clinical responsibility and accountability could not easily be identified or tracked across care transitions from the regional hospital emergency department to the admitting site and specialist department. Missing documentation for the initial ambulance transfer and Royal Flying Doctor Service registration may have included the names of clinicians who registered and received the calls. However, evidence from other patients’ documentation indicated that including this information was not a common practice. A doctor’s name was included on a pre-printed sticker attached to the city ambulance Patient Care record but it was not evident which site or unit this doctor was aligned with. The use of attendant codes in the ambulance service ensured easy tracking of care providers in this phase of the journey and the flight nurse complied with all required identification procedures.

Once at the emergency department, the use of an identification code provided a clear indication of the triage nurse and, while it was apparent from the different initials used that three nurses completed sections of the emergency department documentation, only two completed the signing authority section and the names of the nurses were not legible. When Shirley was seen by the emergency department intern, the notes did not record the time of the consultation, contact details or a legible name. The time of the first medical consult was, however, entered in the electronic patient record. No date, time, contact details or designator details were recorded for the specialist’s entry. However, the medical officer’s name and specialty area was legible. When consultant information was first registered on Shirley’s integrated admission form, three consultants’ names were listed, with no specialty areas provided. This assumes prior knowledge of the clinicians’ expertise and creates potential uncertainty as to who would be overseeing the admission.

Moving to the end of Shirley’s three-day hospital stay and surgery, two discharge letters were held on record for Shirley’s admission. The first, for her transfer to an off-site ward for 11 days of rehabilitation, was completed by an intern on behalf of the orthopedic consultant. This letter was addressed and marked as being sent to a rural general practitioner located 900 km from Shirley’s current residential address. From subsequent documents it was found that this general practitioner no longer managed Shirley’s health care. The second discharge letter back to community care was also completed by an intern on behalf of an orthogeriatric neck of femor specialist at the rehabilitation site. A copy of this second discharge letter was forwarded to Dr D Attending at the doctor’s rooms in Shirley’s home town.

9.4Situation and observations

The communication of a concise appraisal of the situation and provision of supporting clinical findings and observations is important for establishing a shared mental model of the patient’s diagnosis, the stability of their condition and any potential risk of deterioration. The Australian national standard 9 recognizes the essential role that the timely measurement and correct interpretation of physiological observations plays in detecting a patient’s deteriorating condition. The intention of standard 9 is to ensure that any patients who are at risk of deteriorating or who are in decline are recognized promptly and appropriate action is taken to escalate their care. The accurate monitoring and charting of observations provides the opportunity for clinicians to track and review changes in the patient’s condition over time, contextualizing their clinical progress.

In all patient expect records information relating to the clinical situation was communicated in some form. In three instances an abbreviated preliminary diagnosis was the only notation in the record (e.g. #NOF, NSTEMI, and PR BLEED). In a further 35 cases (27.5%) a similarly brief diagnosis was the only clinically related information recorded. The additional information was about the sending site or receiving unit as shown in example 1 below. For the remaining cases additional information included brief descriptions of onset (31%), patient stability (20%), clinical findings (43%) and specific observations (13%). In several instances the provisional diagnosis and either a single (example 2) or a list of vital signs (example 3) were provided.

In the patient expect calls 42% of records (N = 54) made a general reference to the patient’s condition or stability with comments such as HD STABLE, ABCOK, AFEB-RILE, NV STABLE or CLINICALLY WELL. For 9 of the 20 trauma and neurosurgery patients a rating on the Glasgow Coma Scale was provided (e.g. GCS = 5). Specific observations were communicated for 16 patients. However, in the majority of these cases the information provided was either only one piece of information relating to either temperature, heart rate or blood pressure (example 4 below), or part of the information was omitted through use of a noncompliant reporting format (examples 5 and 8). Communication of specific observations to substantiate a general interpretation (as seen in example 6) was not common. Reference to any change in a patient’s condition was made in only three patient expect calls (examples 7 and 8). However, in these cases no specific clinical details were provided to explain the nature or extent of the patient’s deterioration. As a result the broader clinical context and relevance of the change was not made clear.

In the ECCHo survey of clinicians (discussed in chapter 2), respondents who provided accounts of adverse events specifically relating to handover indicated that incorrect communication of patient observations had often been a major contributing factor in the incident. Accounts referred to instances when a patient’s condition was handed over as normal when their observations were in fact within MET call ranges. Unfortunately, several of these cases resulted in the patient’s death. The causes underlying such events are multifaceted, with possible factors a lack of clinical knowledge, incorrect clinical assessment and inaccurate, ambiguous or insufficient communications. However, such mistakes can also be associated with perceptual misjudgement linked to poorly designed charts and ineffective recording systems that rely on one-off or short-term sets of clinical data rather than the use of temporally contextualized formats that provide clear evaluation criteria for identifying a patient’s deteriorating condition. Examples of non-compliance with procedures for recording, inaccuracies in completing charts, poor legibility and design deficiencies were apparent in the written records contained in the patient files audited for this study.

In the Situation and Observations sections of 54 of the records, references were made to clinical findings about tests and scans conducted at the sending site. This information can as easily be communicated under the Background section of iSoBAR. Two examples of clinical findings are provided in examples 9 and 10 below. In both instances the repeated results for a troponin test and reference to a time frame were provided. These examples raise two issues about compliance with standards and the potential for ambiguity. Protocols for high sensitivity troponin T clinical testing state that repeat tests should be conducted after an initial assay (Thygesen et al. 2012). The wording in example 9 allows the reader the opportunity to clearly track the timing of events that confirm a slightly elevated reading. In example 10 the temporal sequencing of test results is unclear. Although the results are within normal limits it is impossible to determine whether the levels are increasing or decreasing over time.

In addition, the use of recording metrics for the troponin tests in these examples and other audited records were also inconsistent. To reduce confusion in reporting conventions, international guidelines and local pathological services recommend that new generation troponin results be reported as whole numbers (old units multiplied by 1000) that is as 15 ng/ L rather than 0.15 (Thygesen et al. 2012). Example 10 uses this whole number metric, implying that a new assay was being used. However, example 9 uses old units. As both patients were processed through the same rural emergency department within a relatively close time frame it is difficult to determine if the clinical findings are from different tests (example 9 from an old assay and example 10 from a new high sensitivity assay) or whether both findings are derived from the same test but an error in metric usage has occurred (i.e. noncompliance with the reporting protocol).

Shirley’s case provides a unique opportunity to consider how such issues were managed for one rural transfer patient. The description of Shirley’s situation in the patient expect call is brief (R # NOF – SUBCAPITAL NV INTACT). Subsequent notations in the triage electronic comments (INJURY – RIGHT LOWER LIMB-? FRACTURE – CLOSED NOF- CVS STABLE- REQUIRING ANALGESIA ENROUTE) and emergency department diagnosis (+ INJURY -+ FRACTURES/CLOSED -+LOWER LIMB -+ FEMUR-OTHER PART OF NECK OF FEMUR) both indicate the possibility of additional lower limb fractures although these are not confirmed in later notations. The side of injury location was noted and clearly drawn on the ambulance documentation but was not included in the Royal Flying Doctor notes, on body diagrams on the triage assessment or in the emergency department intern’s written notes.

Observations noted on Shirley’s flight documentation in the ‘present condition’ section included the initial A.B.C.D assessment (Airway, Breathing, Circulation, Decreased Consciousness). Charted observations of vital signs included temperature, pulse, respiratory rate, oxygen saturation, blood pressure, a pain score and neurovascular observations. The time was also noted when all vital observations were taken during the flight. Intervals fluctuated between 25, 15 and 30 minutes. The flight nurse noted that Shirley had an UNEVENTFUL FLIGHT and was STABLE ­ THROUGHOUT. The ambulance officers also noted A.B.C.D.E (E indicating Everything else) information and that the FLIGHT NURSE AGGRAVATED INJURY DURING TRANSFER. This acknowledgement of aggravation to Shirley’s injury was not noted on the Royal Flying Doctor Service documentation or any subsequent documents. Shirley’s respiratory rate and blood pressure both increased slightly during the road transfer, from 10 inflight to 18 and from 131/54 to 160/?, respectively. Ambulance officers took observations at 15 minute intervals, including Glasgow Coma Scale and pain scores. Shirley’s pain score had reduced to zero in the ambulance with the administration of morphine after landing. Importantly, due to the use of separate forms by each service provider there was no capacity for clinicians to conduct continuous charting of the patient’s observations across the patient journey or to subsequently provide one comprehensive record of this to the receiving clinicians.

In sum, providing a concise assessment of the patient’s diagnosis sets the basic context from which clinicians can establish a shared mental model of the patient’s current situation and ongoing needs. The selection of which clinical findings and observations are relevant for each patient highlights the need for flexible standards when formulating the minimum datasets required for handover communications in different specialist medical areas. This touches on the need to find a balance between content specificity and generality. Evidence from the patient expect calls and written documentation indicate that, while handover content is purpose and function driven, there are obvious communalities. Our data also indicates that consideration should be given to the accessibility and relevance of clinical information to all health professionals working with the patient. For handovers conducted without access to the patient’s observation charts, efforts should be made to highlight the extent and type of change in the patient’s condition leading up to the handover point and any perceived risk of deterioration during the time frame for transfer. The provision of specific clinical biomarkers to substantiate subjective interpretations of the patient’s stability or deterioration is a potential failure detection mechanism for clinicians, providing the opportunity for others to re-evaluate objective information and confirm clinical judgments.

While the Situation and Observations sections of handover focus on articulating the patient’s presenting problem, the broader context of the patient’s pre-existing conditions and relevant past medication history should also be considered when determining and communicating a comprehensive care plan at handover. These elements are discussed under the Background section of iSoBAR.

9.5Background

Just as the situation and observations sections of the iSoBAR handover structure are linked to recognizing and responding to clinical deterioration in patients in acute care (Australian national standard 9), the Background section can be used to incorporate core clinical issues identified in a variety of the Australian national standards into a minimum dataset for handover. These issues include:

–Preventing and controlling health care associated infections (standard 3)

–Ensuring medication safety (standard 4)

–Handling blood and blood products (standard 7)

–Preventing and managing pressure injuries (standard 8)

–Preventing falls and harm from falls (standard 10).

When handing over a patient’s background, information can be logically categorized under two subheadings: pre-admission history and post-admission activity. The intention of this section of the iSoBAR handover protocol is to provide all health care professionals with sufficient relevant background information to help them evaluate and plan the patient’s ongoing care. A core aspect of this section is communication of past medical and social history, particularly co-morbidities that could affect the patient’s current circumstances. Aspects to be considered include the patient’s psychosocial functioning, mobility, language issues, behavioral issues and mental health status. Details of existing medications, known allergies, micro-alerts and identified risk factors such as smoking, alcohol and substance use can also be communicated as and if relevant. The effective communication of medication information and, by association, medication errors is identified in Australian national standard 4 as a major concern in Australian health care and is discussed in greater detail below.

The patient’s background can highlight the results of screening tests to identify patients who pose infection hazards or who are at risk of experiencing pressure injuries and falls and can indicate strategies employed to manage these situations. As circumstances change the prioritization of any background issue can be moved up to the current situation status or discussed in terms of anticipatory guidance during the planning of clinical care (Flemming & Hüber 2013). For example, if a patient like Shirley with COPD and a history of osteoarthritis is identified through screening as being a high falls risks, this risk might be discussed as background information to justify plans for preventative nursing care strategies. The identification of postadmission orders for tests, procedures and consultations that have been undertaken or results that are still pending would also fit under this background section. Such information about what has been done would help shape and inform decision-making about what still needs to happen. As the patient’s journey nears completion, activities of daily living (ADLs), living conditions, social support and medication management are likely to become more relevant for effective discharge planning.

Given the instigation point of patient transfers in our study (from a rural emergency department site) and the purpose of the patient expect call, background information was surprisingly limited. Fewer than 50% of cases provided any form of background information on patients, as shown in example 11 below. Background details most commonly recorded in the patient expect call were references to past medical history, including an acknowledgement of no known history (N = 29, 23%, example 12), and/or treatment currently being administered, including pain relief (N = 24, 19%, examples 13 and 14). Reference to medications history occurred in only five records (4%). No records included specific drug allergies and an acknowledgement of no known allergies (NKA) was found in only two patient expect records. Post-admission information such as the insertion of an indwelling catheter (+ IDC), intravenous line access (IV) and nasogastric intubation (NGT) was communicated in 28% of calls.

The abbreviation for no known allergies (NKA) and variations of no significant medical history (NIL SIG HX) were used repeatedly in the written documents and provide an expedient way to indicate that a patient’s history had been taken into consideration. However, the absence of reference to a patient’s medical history did not mean that a patient’s history was not a significant factor in the admission. For example, in only one of the 127 patient expect calls was reference to a micro alert communicated for a patient (example 15). However, information extracted from the audit of written documents indicated that both infection control alerts and severe medication allergies had been known for several patients but had not been communicated during the patient expect call. In the 12 fully audited patient files one patient was identified with a micro alert category (vancomycn resistant enterococci, VRE). A further three patients required infection control precautions on admission: two patients for MSRA identified risk required white level precautions (separate room) and one patient required contact (orange) and droplet (green) level transmission-based precautions. In none of these cases was the micro alert information communicated in the patient expect call. Again this type of omission during handover was raised as a serious concern for workers in the ECCHo handover survey, as were medication related errors.

Medication errors have been defined as ‘a failure in the treatment process that leads to, or has the potential to lead to, harm to the patient’ (Aronson 2009, p. 601). Ferner and Aronson (2006) describe how medication errors involve an array of knowledge-, skill-, action- and memory-based errors that can occur across the full treatment process. These range from mistakes made during the decision-making processes when prescribing medications, to skill-based errors (slips and lapses) in both writing prescriptions and administering medications. Our focus here is on identifying common errors in the communication and documentation of the patient’s medication history that may contribute to poor patient outcomes.

Past research, referred to in the Australian national standard 4 on medication safety, has indicated that one or more errors can be found in up to 70% of medication histories; nearly half of all medication errors occur at transitions of care; and approximately one-third of medication errors have the potential to cause harm (Tam et al. 2005; Cornish et al. 2005). To ensure the continuity of patient medication information, Australian national standard 4 recommends that:

While the inclusion of a complete medication list during the patient expect call could be considered an unrealistic expectation, some reference to medications was made in 19% of the calls. These were largely in reference to the use of beta blockers, inotropes, dopamine, warfarin, IV antibiotics and medications administered for pain relief prior to transfer. As stated above no patient expect calls included information on specific drug allergies, although evidence from the 12 audited patient files showed that five patients had known drug allergies, some of which were severe. Our review of both the electronic records and patient files showed:

As an example of these issues, Tabs. 9.6 and 9.7 summarize the information recorded in documents for Shirley’s rural transfer. Of particular concern was the timing of efforts made to clarify Shirley’s medication and allergy history. A request to Shirley’s local general practitioner rooms to provide a comprehensive list of existing medications and allergies was made by clinicians at the rehabilitation facility Shirley was discharged to seven days after her initial rural transfer and admission for surgery. This list of five medications and the associated nurse’s comments are labeled Med Summary in Tab. 9.6 and Fax 1 and 2 in Tab. 9.7. This list differs from Shirley’s transfer and admission records, which only record her allergies to metronidazole, amoxycillin and lisinopril. The convention to use lower case letters for generic drug names and a capital letter for drug brand names, as used here, was not followed consistently in the documents. The medical summary showed that Shirley had previously had an anaphylactic reaction to the broad spectrum antibiotic cephalexin in 2009 and also listed the drug class cephalosporins in the allergy summary. Neither the specific drug nor class were included in the transfer and admission handover records. Allergy notations by the triage nurse, emergency department nurse, emergency department intern and reviewing orthopedic registrar in the emergency department were inconsistent, typically identifying only two or three of Shirley’s allergies, and either omitting or interchanging lisinopril and ACE inhibitor class drugs in the list.

Tab. 9.6: Medication lists and variations across a patient journey

Note: Miss = Document missing from file; @ = Purpose of medication provided; # = Patient denies use of medication; ED IN = emergency department Intern Notes; ED Orth = emergency department Orthopedic Notes; ? = query purpose
*Medication summary from patient’s regular health provider was sought by the metro rehabilitation facility after Discharge 1 to Rehabilitation Unit Tab.

Tab. 9.7: Allergy lists and variations across a patient journey

Note: Miss = Document missing from file; @ = Reaction listed; % = No History of side effects; # Augmentin is an oral antibacterial combination consisting of amoxicillin and the beta-lactamase inhibitor, clavulanate
*Medication summary from patient’s regular health provider was sort by the metro rehabilitation facility at stage 11 & 12.

In the final discharge documents for Shirley, her drug allergies are listed as ‘metronidazole, ACE inhibitors, amoxycillin, cepfalexin, lisinopril and Augmentin’. There are several issues with this final list. First, there is a discrepancy between the listing of amoxicillin as a drug allergy in the transfer and triage patient records and the record of Amoxil 500 mg capsules (t.i.d. – three times per day) for an unspecified infection on the general practitioner medication summary. The notation in the faxed medical summary did identify Shirley’s reaction to Augmentin, which is an oral antibacterial combination consisting of amoxicillin and the beta-lactamase inhibitor clavulanate. Identification of amoxicillin as a new allergy was not made clear in any patient records and this conflicting information did not appear to be adequately resolved at discharge.

A second issue was a notation on the faxed general practitioner medication summary stating that there was ‘no history of side effects from Flagyl or Doxycycline’. This statement did not clarify whether Shirley in fact had an allergy to metronidazole (Flagyl) or that her records simply did not included the side effects she experienced. Possibly as a precautionary action both metronidazole and amoxicillin were retained on the discharge list of allergies. Finally the discharge list included the drug class ACE inhibitors (which includes lisinopril) but dropped cephalosporins (the drug class which includes cephalexin). While the intern completing the discharge summary appears to have taken a conservative approach to the inclusion of some drugs – if in doubt keep them in – the omission of the drug class to which Shirley has shown the most severe reaction seems aberrant.

In the documentation of Shirley’s medications, it was also difficult to ascertain if prescribed medications were unintentionally omitted or intentionally ceased as no justifications of the medications listed or changes made were provided. Discrepancies in medications involved the omission from the ambulance, Royal Flying Doctor and emergency department records of Shirley’s daily Caduet tablet for ischemic heart disease. The daily use of sertraline, for depression, was also listed in the general practitioner summary and most records although no mention of ischemic heart disease or depression was made in Shirley’s medical history. The general practitioner medical summary also indicated that Shirley was prescribed the steroid prednisolone (30 mg tablets taken daily) and this was listed in ambulance and triage notes. The emergency department intern makes an illegible notation regarding prednisolone in the emergency department plan, and in the emergency department orthopedic review the clinician notes that Shirley ‘denies regular Prednisolone’ use. Prednisolone is not listed again in the patient records or discharge summaries and a cessation of its use is assumed. The use of temazepam for insomnia is also ceased on the discharge letter from the rehabilitation ward. The purpose of each new and existing medication was only explicitly stated in Shirley’s discharge documents. In total, Shirley’s original medication summary indicated that she was prescribed 19 supplements, inhalants and oral medications. On discharge this was reduced to 11, the major changes being the cessation of supplements and the prescribing of risedronate and calcium supplements for osteoporosis.

9.6Agreed plan

The Agreed Plan/ Actions sections of the iSoBAR protocol is where clinicians should stipulate what actions should be undertaken and why. For rural transfer patients the details of plans varied considerably across the different documentation contexts. In the patient expect call, clinicians provided brief information about both logistical and clinical plans. In the transfer agency records, logistical details (i.e. where to, who to and how to) took precedence over treatment plans (i.e. patient needs, flight medications and observations timing). In the hospital setting, the focus shifted to medical assessment and treatment. A noticeable omission across all documentation was any statement of the rationale or explanation for decisions taken or orders made.

In 53% of the patient expect calls some form of planning information was included. In terms of logistics, a notation that the patient would be arriving via Royal Flying Doctor Service was made in 32 (25%) cases. Comments about the admission process, including instructions to call the admitting ward/clinician occurred in 22 (17%) instances (example 16 below). For forward planning purposes, the patient’s expected arrival time was recorded under a separate heading in the electronic record. When an expected arrival time was not available this was generally noted in the patient expect record with comments such as UNCLEAR AT THIS POINT COMING Royal Flying Doctor Service WHEN ABLE; NO RESOURCES TO TRANSFER AT THIS STAGE; WILL FLY WHEN WEATHER PERMITS. Arrival times were updated at a later point in such cases. The average delay between expected and actual arrival time for rural patients was 195 minutes. Twenty-five percent of the patients arrived within 22 minutes of the arrival time stipulated in the patient expect call; 50% within 83 minutes and 25% longer than four hours. Information about actions to be taken after the patient’s arrival was included in 24 (19%) calls. This planning information was restricted to simple instructions: who to contact, review instructions (examples 17 and 18), and – in fewer cases – tests and procedures recommended or scheduled (examples 19 and 20).

9.7Readback

The Readback component of iSoBAR relates to the opportunity to access and clarify information. Any documentation should provide readers with the opportunity to track who was responsible for what decisions and actions, when they occurred and why they were needed. The importance of ensuring the ‘traceability of decisions made to enable understanding [of] the case’ is highlighted by Flemming and Hübner (2013: 587) in their discussion of the need for clinicians to establish common ground in handover practice. Having high quality technology support for handover makes common ground observable and shared knowledge accessible (Patterson 2012). If supporting documentation is convoluted, illegible, incomplete, inaccurate or ambiguous the opportunity for other health workers to refer back to clarify or confirm specific details in the paperwork is impaired. Just as poor quality of verbal interactions can limit a handover receiver’s capacity to ‘make sense of the patient’s unfolding episode of illness and treatment and to take appropriate actions’ (Cohen et al. 2012: 304), so too do poor quality written interactions impact on a reader’s ability to create an accurate shared mental model of the patient’s progress.

Readback is, in the literal sense, the capacity for patient-related information to be read back effectively. The written documents are therefore the accessible and enduring resources that support the sequence of distinct, verbal handover events. In terms of designing a resilient system that optimizes the detection and recovery from errors, the capacity to read back the clinical notes to clarify ambiguities and identify inconsistencies is itself a fundamental ‘failure detection mechanism’ (Thomas et al. 2013: p. 51). To flesh out how the capacity for readback can be improved in handover, the following section describes issues in the patient expect call and core considerations for written communications involving C-A-R-E: compliance, accessibility, readability and the enduring capacity of such records to explain the patient’s experience.

9.7.1Compliance

As a starting point, the electronic and written documents used to support handover need to comply with the standards and procedures of the health services and professional bodies concerned. Health professional therefore need to be aware of the regulatory and professional standards required when completing documentation. While this point should be automatic and irrefutable, the evidence from our audit indicates that noncompliance is commonplace, due to mistakes and routine violations of procedures. If health care workers are routinely violating or working around procedures when producing or accessing documentation, this should be identified as a red flag. It indicates ineffective procedures or underlying system failures that should be raised with governing bodies for consideration and action, enhancing the organization’s capacity to learn from failure (Dekker 2011).

9.7.2Accessibility

Under accessibility we consider two components: tangible access and interpretative access. Tangible access is the requirement that the information be physically available to all who are authorized to access to it. Interpretative access is the requirement that information be provided in a manner that is easily understood by the intended audience. The first is a consequence of good information management systems; the second relates to good communicative process and relies on the writer understanding the full purpose of the notation and respecting the capacities of the potential audience for whom the document is intended.

9.7.3Readability

Readability concerns the more prosaic and practical aspects of the written documents. It includes the legibility of the information (typically challenging) and the use of grammar, abbreviations and symbols. Readability also involves the application of technologies and design issues. While the shift to electronic mediums may reduce the challenges imposed by illegible handwriting in patient medical records and charts, the condensed forms of communications in many electronic records pose an entirely different set of problems. A form or chart that is poorly designed – one that does not provide sufficient space to record the information required, does not provide sufficiently clear instructions for completion or includes redundant sections – is ineffective and inefficient. Similarly, an electronic record system that does not provide an intuitive interface may be problematic for readability.

One of the main issues in the patient expect calls was the inconsistency and over-use of abbreviations. For example, nonclinical abbreviations were applied inconsistently in the patient expect notations with EX being used in some instances to represent the term expected by (as in example 21 below) or more commonly to represent from this site (as in example 22). Examples of clinical abbreviations that were used inconsistently in the records include HD, PR and PE. HD (as in example 23 below) relates to the patient’s condition being hemodynamically stable but elsewhere in the data HD referred to hemodialysis and Hodgkin’s disease. PE can be used to represent pulmonary embolism, pulmonary edema, pleural effusion and pericardial effusion. In electronic records the typed text is generally far more legible in form than handwritten records, although typographic errors and prolific use of abbreviations affected readability. In a number of our audited records the entire communication for the patient expect call was in either truncated, abbreviated or symbolic form as in examples 21 to 24, as follows:

Abbreviations were also often used for headings within the notations. The use of headings or signposts to set the context of subsequent information can be a useful communication strategy to improve readability, but problems can arise when inconsistent or ambiguous abbreviations are used. For example, in the patient expects calls, headings were used to distinguish between the patient’s current medical situation and their past or coexisting medical conditions. However, past medical history was inconsistently shortened to PMHx, PMH, PMx, Hx, background of and known. While the meaning of the term remains clear in the majority of cases, the potential for ambiguity is illustrated in the following notation:

In the local context, young adults suffering from mental health disorders may continue their treatment in the local children’s hospital. This particular local facility is known by the acronym PMH. The combination of the words and abbreviations used in the notation is potentially ambiguous and could imply that the patient may require a subsequent transfer to the other facility for ongoing treatment of their pre-existing condition after their presenting orthopedic admission is managed.

9.7.4Endurability

This last component highlights the need for clinicians to provide in the documentation an enduring, explanatory record of the patient’s treatment episode. Bringing together all the other components, the records should have the capacity to be interpreted beyond the immediate time and situational context. To ensure the integrity of the information endures, clinicians need to consider providing explanations of why decisions are made and actions taken. To improve efficiency, health services should also consider developing written or electronic documentation systems that support confirmatory reviews of existing entries rather than duplication of entries. For example, duplication of information was apparent in the clinicians’ and specialists’ notes recorded in the emergency department. In this instance it appeared that each clinician started the clinical information gathering process from scratch, implementing their unique interview and recording practices. While the vast majority of the information generated was subsequently common, any inconsistencies, inaccuracies or omissions between the two records were rarely acknowledged or reconciled. However, if processes and forms were redesigned to allow space under standardized headings for a second clinician or specialist to confirm the existing clinical information recorded in the initial patient interview with the patient and then to correct and expand this information with relevant details when and where necessary, a functional shift from information generation to confirmation would be supported.

When inconsistencies or inaccuracies are identified through confirmatory review processes, providing a justification of any changes or omissions would improve the clarity and continuity of information. Communicating the purpose of any past or planned actions – rather than simply identifying tasks – potentially supports the development of a more comprehensive shared understanding of the patient’s needs with other health workers accessing the documents.

A lack of compliance with standards for identification and accepted medical abbreviations can be classified as routine violations of procedures. In relation to the accessibility and readability of the electronic record it is important to note that the expected audience for the patient expect call was the knowledgeable and interconnected group of clinicians operating in the metropolitan emergency department. The assumption of an audience familiar with each other and conversant with local communication conventions is reflected in the minimal use of clinician names and/ or initials as identifiers and the prevalence of abbreviations. In numerous instances both common usage and medical abbreviations were ambiguous outside the immediate clinical context, their interpretation being dependent on readers having prior awareness of the medical specialty with which it is associated. Both these aspects of ambiguity place in question the enduring nature of the communication.

To focus on Shirley’s interhospital transfer experience, documentation for several stages of the transfer process were missing from Shirley’s audited patient records, bringing into question the long-term accessibility of transfer-related information. In the documents that were available we observed a series of skill, rule and knowledge-based errors and evidence of routine violations (Rasmussen 1982; Reason 1990). Examples included inconsistencies in medications and allergies, incorrect patient descriptors, inconsistent and inaccurate use of symbols and abbreviations and noncompliance with standard reporting protocols. In terms of the handover of clinical responsibility and accountability for Shirley’s care in these initial stages of her patient journey, the system did not result in any explicit adverse events. However, the errors associated with Shirley’s allergies and admission experience could justify labeling it a ‘near miss’ event.

9.8Summary: expanding the concept of written clinical communication

From our audit of written documents, we identified numerous problems that had clear potential to jeopardize patient safety and that infringed national safety and quality standards. Above we have outlined specific recommendations to address the most serious of these problems. As we will now explain, our findings also led directly to the iC3 handover communication model explained in detail in chapter 15.

In order to formalize and generalize our findings, we built on the communication strategies relating to verbal handover described by Eggins and Slade in earlier chapters and in Eggins and Slade (2012 and 2016). Eggins and Slade’s (2012) conceptualization of communication distinguishes between what information is communicated (informational strategies) and how it is communicated (interactional strategies). In their work on spoken bedside handovers (chapter 6) Eggins and Slade formulated the CARE acronym (Connect, Ask, Respond, Empathize) to capture the interactional strategies required in an effective spoken handover. To specify effective informational strategies in spoken handovers, they adopted and adapted the locally required iSBAR protocol (chapter 7).

In light of our interhospital transfer audit research, we began by differentiating between the informational and interactional dimensions of handover communication We focused, firstly, on the informational structure used for selecting and organizing relevant patient information. Informational structure refers to both the selection of specific information to be conveyed (the minimum dataset) and the standardized sequence of stages in which that information is to be presented. Informational structure is largely dependent on compliance with the enacted state or organizational policies and procedures and the specific clinical context. In our context – and, increasingly, across Australia – iS(o)BAR is the informational structure required, although this is a flexible component which can as easily incorporate other site-specific standardized handover formats (e.g. SBAR, SBARR, iSBAR, De MIST). Table 9.8 displays the iSoBAR category labels we settled on, as discussed earlier in this chapter.

Information structure provides the categories and ideal staging or sequencing of information in handover. However, as our research showed, the quality of the information handed over in that staged sequence is equally critical. For example, what matters is not just that a clinician provides some Situation and Background, but that the information in those stages is what is needed to ensure the patient’s safe and continuous care. In other words, clinicians need to apply effective informational strategies. As a second step, then, we focused on the information strategies clinicians need to achieve effective written handover.

Tab. 9.8: Informational structure in handover captured in the iSoBAR protocol

Tab. 9.9: CARE interpreted as an informational strategy

To arrive at explicit informational strategies for written information, we found it useful to re-interpret Eggins and Slade’s CARE acronym. As discussed in this chapter, our audit showed that written information delivered in handover needs to be Concise, Accurate, Reasoned and Explicit. We therefore added this element of informational strategy to the model, as in Tab. 9.9.

A third step we took was to consider the interactional dimension of written communication. As we tracked the transfer of information across hospital sites we realized that the interactional strategies Eggins and Slade identified as crucial in spoken handovers applied equally to written handovers. That is, as Eggins & Slade (2012, 2016) point out, to gather and share the relevant patient information clinicians need to develop effective interactional strategies with patients and fellow health workers across communication media. These interactions include both verbal communications – conducted face-to-face, via telephone or videoconference – and, as our audit of interhospital transfer showed, written communications, such as patient’s notes, emails, letters and computer records. While all these media differ, to be effective they need to be interactionally sensitive – that is, designed and delivered to meet the needs of all the participants involved. These include the need to show an understanding of and respect for all participants in the communication, whether by empathizing with a patient in pain when being interviewed, by asking the incoming clinician’s opinion about a patient’s care plan at handover, or by ensuring patient notes are unambiguous and legible at the completion of the shift.

To capture the requirements of effective interaction through written communication, we again found the CARE acronym useful. We reinterpreted CARE to give a set of interactional strategies for written clinical communications supporting handover – see Tab. 9.10 below. Ensuring the ongoing continuity and accuracy of all patient information and records of health professional actions is imperative for optimal patient care and safety. The priority strategies for written communications considered here have been compliance with standards and procedures; accessibility to audience, readability and the enduring nature of the records as an explanation of the patient’s journey.

Tab. 9.10: Adapting CARE to capture interactional strategies in written handover

Interactional Strategy (Written) Make sure that everything you write is:
Compliant with policies & standards	Compliant	C
Accessible to your audience	Accessible	A
Legible	Readable	R
Durable and self-explanatory	Enduring	E

The final iC3 model is presented in chapter 15, where it is also discussed in detail. It is comprehensive, empirically derived and theoretically motivated. It applies to both written and spoken handovers and captures both the informational and interactional dimensions of communication. The model allows us to better understand the risks that may arise across the handover process and the strategies clinicians can deploy to minimize those risks. In Shirley’s case, had the informational and interactional strategies in the iC3 model been adhered to more rigorously, the vulnerabilities that had the potential to cause her harm would have been reduced if not eliminated.