Iterative Development of Clinician Guides to Support Deprescribing Decisions and Communication for Older Patients in Hospital: A Novel Methodology.

BACKGROUND/OBJECTIVES: Medication review is an important component of the management of older hospital patients. Deprescribing (supervised withdrawal of inappropriate medicines) is one outcome of review. This study aimed to iteratively develop and test the usability of deprescribing guides, which support multidisciplinary clinicians to reduce inappropriate polypharmacy in older inpatients. METHODS: Deprescribing guides for hospital clinicians were developed using a novel mixed-methods, ten-step process. Iterative development and usability testing were applied. This included content development through review of the literature; expert consensus through five rounds of feedback using a modified Delphi approach; and usability testing by 16 multidisciplinary hospital clinicians on hypothetical clinical scenarios involving observations, semi-structured interviews, and administration of the System Usability Scale. RESULTS: This novel process was used to develop deprescribing guides that facilitate implementation of evidence on deprescribing in routine hospital care. The guides present evidence-based information in a format that aligns with workflows of multidisciplinary hospital clinicians. The guides were adapted for various clinical roles to navigate efficiently to suit differing workflow needs. Guides include unique communication support in the form of "preferred language". Clinicians can use the "preferred language" to apply the evidence to the individual patient and relay decisions between health providers and with patients/carers. The total System Usability Scale score was 80.6 ± 2.0 (mean ± standard error of the mean), indicating excellent usability. Guides have been developed using consistent format for nine medication classes that are common targets for deprescribing and are publicly available. CONCLUSION: This study demonstrates a novel approach to the development and implementation of evidence-based recommendations that support deprescribing in routine hospital care.

Evaluation of Clinical Relevance of Drug-Drug Interaction Alerts Prior to Implementation.

INTRODUCTION: Drug-drug interaction (DDI) alerts are often implemented in the hospital computerized provider order entry (CPOE) systems with limited evaluation. This increases the risk of prescribers experiencing too many irrelevant alerts, resulting in alert fatigue. In this study, we aimed to evaluate clinical relevance of alerts prior to implementation in CPOE using two common approaches: compendia and expert panel review. METHODS: After generating a list of hypothetical DDI alerts, that is, alerts that would have been triggered if DDI alerts were operational in the CPOE, we calculated the agreement between multiple drug interaction compendia with regards to the severity of these alerts. A subset of DDI alerts (n = 13), with associated patient information, were presented to an expert panel to reach a consensus on whether each alert should be included in the CPOE. RESULTS: There was poor agreement between compendia in their classifications of DDI severity (Krippendorff's α: 0.03; 95% confidence interval: -0.07 to 0.14). Only 10% of DDI alerts were classed as severe by all compendia. On the other hand, the panel reached consensus on 12 of the 13 alerts that were presented to them regarding whether they should be included in the CPOE. CONCLUSION: Using an expert panel and allowing them to discuss their views openly likely resulted in high agreement on what alerts should be included in a CPOE system. Presenting alerts in the context of patient cases allowed panelists to identify the conditions under which alerts were clinically relevant. The poor agreement between compendia suggests that this methodology may not be ideal for the evaluation of DDI alerts. Performing preimplementation review of DDI alerts before they are enabled provides an opportunity to minimize the risk of alert fatigue before prescribers are exposed to false-positive alerts.

Drug-drug interactions and their harmful effects in hospitalised patients: a systematic review and meta-analysis.

PURPOSE: Drug-drug interactions (DDIs) are often avoidable and, if undetected, can lead to patient harm. This review aimed to determine the prevalence of potential DDIs (pDDIs), clinically relevant DDIs (DDIs that could lead to measurable patient harm, taking into account the patient's individual clinical profile) and DDIs that resulted in actual patient harm during hospitalisation. METHOD: Four databases were scanned for English papers published from 2000 to 2016. Papers that reported prevalence of DDIs in the outpatient setting, at admission or discharge, involving only specific drugs, or in specific disease populations or age groups were excluded. RESULTS: Twenty-seven papers met the inclusion criteria and were graded for quality using the Critical Appraisal Skills Programme (CASP) cohort study checklist. Ten papers were rated as 'poor', 14 as 'fair' and only three papers as 'good'. Overall, the meta-analysis revealed that 33% of general patients and 67% of intensive care patients experienced a pDDI during their hospital stay. It was not possible to determine the prevalence of clinically relevant DDIs or DDIs that resulted in actual patient harm as data on these categories were limited. Of the very few studies that reported on harm, only a small proportion of DDIs were found to have resulted in actual patient harm. CONCLUSIONS: Standardisation of DDI definitions and research methods are required to allow meaningful prevalence rates to be obtained and compared. Studies that go further than measuring pDDIs are critically needed to determine the impact of DDIs on patient safety.

A systematic review of the effectiveness of interruptive medication prescribing alerts in hospital CPOE systems to change prescriber behavior and improve patient safety.

OBJECTIVES: To assess the evidence of the effectiveness of different categories of interruptive medication prescribing alerts to change prescriber behavior and/or improve patient outcomes in hospital computerized provider order entry (CPOE) systems. METHODS: PubMed, Embase, CINAHL and the Cochrane Library were searched for relevant articles published between January 2000 and February 2016. Studies were included if they compared the outcomes of automatic, interruptive medication prescribing alert/s to a control/comparison group to determine alert effectiveness. RESULTS: Twenty-three studies describing 32 alerts classified into 11 alert categories were identified. The most common alert categories studied were drug-condition interaction (n=6), drug-drug interaction alerts (n=6) and corollary order alerts (n=6). All 23 papers investigated the effect of the intervention alert on at least one outcome measure of prescriber behavior. Just over half of the studies (53%, n=17) reported a statistically significant beneficial effect from the intervention alert; 34% (n=11) reported no statistically significant effect, and 6% (n=2) reported a significant detrimental effect. Two studies also evaluated the effect of alerts on patient outcome measures; neither finding that patient outcomes significantly improved following alert implementation (6%, n=2). The greatest volume of evidence relates to three alert categories: drug-condition, drug-drug and corollary order alerts. Of these, drug-condition alerts had the greatest number of studies reporting positive effects (five out of six studies). Only two of six studies of drug-drug interaction and one of six of corollary alerts reported positive benefits. DISCUSSION AND CONCLUSION: The current evidence-base does not show a clear indication that particular categories of alerts are more effective than others. While the majority of alert categories were shown to improve outcomes in some studies, there were also many cases where outcomes did not improve. This lack of evidence hinders decisions about the amount and type of decision support that should be integrated into CPOE systems to increase safety while reducing the risk of alert fatigue. Virtually no studies have sought to investigate the impact on changes to prescriber behavior and outcomes overall when alerts from multiple categories are incorporated within the same system.

Stepped-wedge cluster randomised controlled trial to assess the effectiveness of an electronic medication management system to reduce medication errors, adverse drug events and average length of stay at two paediatric hospitals: a study protocol.

INTRODUCTION: Medication errors are the most frequent cause of preventable harm in hospitals. Medication management in paediatric patients is particularly complex and consequently potential for harms are greater than in adults. Electronic medication management (eMM) systems are heralded as a highly effective intervention to reduce adverse drug events (ADEs), yet internationally evidence of their effectiveness in paediatric populations is limited. This study will assess the effectiveness of an eMM system to reduce medication errors, ADEs and length of stay (LOS). The study will also investigate system impact on clinical work processes. METHODS AND ANALYSIS: A stepped-wedge cluster randomised controlled trial (SWCRCT) will measure changes pre-eMM and post-eMM system implementation in prescribing and medication administration error (MAE) rates, potential and actual ADEs, and average LOS. In stage 1, 8 wards within the first paediatric hospital will be randomised to receive the eMM system 1 week apart. In stage 2, the second paediatric hospital will randomise implementation of a modified eMM and outcomes will be assessed. Prescribing errors will be identified through record reviews, and MAEs through direct observation of nurses and record reviews. Actual and potential severity will be assigned. Outcomes will be assessed at the patient-level using mixed models, taking into account correlation of admissions within wards and multiple admissions for the same patient, with adjustment for potential confounders. Interviews and direct observation of clinicians will investigate the effects of the system on workflow. Data from site 1 will be used to develop improvements in the eMM and implemented at site 2, where the SWCRCT design will be repeated (stage 2). ETHICS AND DISSEMINATION: The research has been approved by the Human Research Ethics Committee of the Sydney Children's Hospitals Network and Macquarie University. Results will be reported through academic journals and seminar and conference presentations. TRIAL REGISTRATION NUMBER: Australian New Zealand Clinical Trials Registry (ANZCTR) 370325.

Trends in metformin utilisation and dose appropriateness in Australia.

PURPOSE: The study aimed to (1) determine the trends in the utilisation of metformin in Australia, (2) determine the appropriateness of metformin dosing in an Australian teaching hospital and (3) gather the opinions of prescribers on the relationship between metformin dose and renal function. METHODS: National prescription data between 1990 and 2012 were accessed. A retrospective audit (2008-2012) of metformin doses and patient renal function (20 % random sample of all in-patients prescribed metformin) was conducted at St Vincent's Hospital (SVH), Sydney. Prescribers of metformin were interviewed (semi-structured; consultants at SVH) or surveyed (Australian endocrinologists) to gather their understanding of metformin dosing in relation to renal function. RESULTS: Metformin utilisation increased fivefold nationally between 1995 and 2012. Metformin tended to be under-dosed in SVH patients with normal renal function (83.5 %) and over-dosed in patients with impaired renal function (estimated glomerular filtration rate (eGFR) <30 mL/min, 50 %). Consultants indicated that metformin doses needed to be reduced in renal impairment. Most endocrinologists (61 %) were comfortable prescribing metformin down to eGFRs around 30 mL/min. CONCLUSION: The use of metformin increased greatly over the period of the study. Metformin is prescribed frequently for patients with eGFR values below the minimal level approved in the product label (60 mL/min). While prescribers expressed their understanding of the need to reduce metformin doses in patients with renal impairment, we found that metformin doses were higher than appropriate in patients with impaired renal function. Metformin may be used safely when renal function is poor provided dosage is appropriately reduced.

Evaluation of hospital-wide computerised decision support in an intensive care unit: an observational study.

We conducted an observational study with interviews in a 12-bed general/neurological intensive care unit (ICU) at a teaching hospital in Sydney, Australia, to determine whether hospital-wide computerised decision support (CDS) embedded in an electronic prescribing system is used and perceived as useful by doctors in an ICU setting. Twenty doctors were shadowed by the observer while on ward rounds (33.6 hours) and non-ward rounds (28 hours) in the ICU. These doctors were also interviewed to explore views of CDS. We found that computerised alerts were triggered frequently in the ICU (n=166, in 59% of orders), less than half of the alerts were read by doctors and only four alerts resulted in a medication order being changed. Pre-written orders were utilised frequently, however reference material was rarely accessed. Interviews with doctors revealed a willingness to use CDS features; however the primary barrier to use was lack of customisation for the ICU setting. Doctors working in the ICU triggered a high number of alerts when prescribing, 40% more alerts than doctors working on general wards of the same hospital. Certain procedures in place in the ICU (e.g. daily microbiology ward rounds) made many alerts redundant in this setting. Lack of customisation for the ICU led to dissatisfaction with CDS and infrequent use of some CDS features.

A time and motion study of junior doctor work patterns on the weekend: a potential contributor to the weekend effect?

BACKGROUND: Patients admitted to hospital on weekends have a greater risk of mortality compared to patients admitted on weekdays. Junior medical officers (JMO) make up the majority of medical staff on weekends. No previous study has quantified JMO work patterns on weekends. AIM: To describe and quantify JMO work patterns on weekends and compare them with patterns previously observed during the week. METHODS: Observational time and motion study of JMO working weekends using the Work Observation Method by Activity Timing (WOMBAT; Australian Institute of Health Innovation, Macquarie University, Sydney, NSW, Australia) software. Descriptive statistics were used to determine the proportion of total observed time spent in tasks. RESULTS: Weekend JMO predominately spent time in indirect care (32.0%), direct care (23.0%) and professional communication (22.1%). JMO spent 20.9% of time multitasking and were interrupted, on average, every 9 min. Weekend JMO spent significantly more time in direct care compared with weekdays (13.0%; P < 0.001) and nights (14.3%; P < 0.001). Weekend JMO spent significantly less time on breaks (8.5%), with less than 1 h in an 11-h shift, compared with JMO during weekdays (16.4%; P = 0.004) and nights (27.6%; P = <0.001). Weekend JMO were interrupted at a higher rate (6.6/h) than on weekdays (rate ratio (RR) 2.9, 95% confidence intervals (CI) 2.6, 3.3) or nights (RR 5.1, 95% CI 4.2, 6.1). Multitasking on weekends (20.9%) was comparable to weekdays (18.9%; P = 0.19) but significantly higher than nights (6.4%; P = <0.001). CONCLUSION: On weekends, JMO had few breaks, were interrupted frequently and engaged in high levels of multitasking. This pattern of JMO work could be a potential contributing factor to the weekend effect in terms of JMO abilities to respond safely and adequately to care demands.

Mobile Applications for Patient-centered Care Coordination: A Review of Human Factors Methods Applied to their Design, Development, and Evaluation.

OBJECTIVES: To examine if human factors methods were applied in the design, development, and evaluation of mobile applications developed to facilitate aspects of patient-centered care coordination. METHODS: We searched MEDLINE and EMBASE (2013-2014) for studies describing the design or the evaluation of a mobile health application that aimed to support patients' active involvement in the coordination of their care. RESULTS: 34 papers met the inclusion criteria. Applications ranged from tools that supported self-management of specific conditions (e.g. asthma) to tools that provided coaching or education. Twelve of the 15 papers describing the design or development of an app reported the use of a human factors approach. The most frequently used methods were interviews and surveys, which often included an exploration of participants' current use of information technology. Sixteen papers described the evaluation of a patient application in practice. All of them adopted a human factors approach, typically an examination of the use of app features and/or surveys or interviews which enquired about patients' views of the effects of using the app on their behaviors (e.g. medication adherence), knowledge, and relationships with healthcare providers. No study in our review assessed the impact of mobile applications on health outcomes. CONCLUSION: The potential of mobile health applications to assist patients to more actively engage in the management of their care has resulted in a large number of applications being developed. Our review showed that human factors approaches are nearly always adopted to some extent in the design, development, and evaluation of mobile applications.

Understanding and improving the use of allopurinol in a teaching hospital.

BACKGROUND: An emphasis on renal function in deciding maintenance doses of allopurinol to prevent allopurinol hypersensitivity has resulted in ineffective prevention of attacks of gout. New therapeutic guidelines for gout have shifted the focus back to titrating maintenance doses to reach a serum uric acid (SUA) concentration target of ≤ 0.36 mmol/L. AIMS: To examine trends in the prescribing of allopurinol in a teaching hospital and their concordance with the new guidelines for gout management, and to explore prescribers' approaches and attitudes to the use of allopurinol. METHODS: An audit was conducted of all inpatients prescribed allopurinol at a teaching hospital between January 2008 and December 2012. Allopurinol dose, SUA, serum creatinine concentrations and estimated glomerular filtration rates were extracted from the hospital databases. Doctors from medical units who regularly prescribed allopurinol were interviewed. RESULTS: The allopurinol dose prescribed in gout patients most commonly was a continuation of the pre-admission dosage. Dosage change during admission was rarely observed. Dosages reflected a consideration of renal function. SUA concentrations were measured in only 21% (n = 269) of gout patients. Prescriber interviews (n = 12) reflected adequate knowledge regarding allopurinol use, but most maintained that the primary care setting was more suitable for the management of dose titration in gout. CONCLUSIONS: SUA concentrations were not routinely measured in the majority of admitted gout patients taking allopurinol. Without SUA measurements and allopurinol dose titration, patients with SUA > 0.36 mmol/L are at increased risk for acute attacks of gout in hospital.

iPad use at the bedside: a tool for engaging patients in care processes during ward rounds?

BACKGROUND: Previous work has examined the impact of technology on information sharing and communication between doctors and patients in general practice consultations, but very few studies have explored this in hospital settings. AIMS: To assess if, and how, senior clinicians use an iPad to share information (e.g. patient test results) with patients during ward rounds and to explore patients' and doctors' experiences of information sharing events. METHODS: Ten senior doctors were shadowed on ward rounds on general wards during interactions with 525 patients over 77.3 h, seven senior doctors were interviewed and 180 patients completed a short survey. RESULTS: Doctors reported that information sharing with patients is critical to the delivery of high-quality healthcare, but were not seen to use the iPad to share information with patients on ward rounds. Patients did not think the iPad had impacted on their engagement with doctors on rounds. Ward rounds were observed to follow set routines and patient interactions were brief. CONCLUSIONS: Although the iPad potentially creates new opportunities for information sharing and patient engagement, the ward round may not present the most appropriate context for this to be done.

Shedding light on junior doctors' work practices after hours.

BACKGROUND: It is imperative to understand the current work practices of hospital personnel to inform efforts and secure resources towards the improvement of hospital systems. Research examining doctors' work during night-shifts is limited. AIM: To describe and quantify the night-shift work practices of junior doctors. METHODS: An observational time and motion study was conducted. Eight resident doctors in four general wards were observed for 96 h during night shifts (Monday-Friday, 2200-0800). RESULTS: Doctors spent the highest proportion (28%; 95% CI 21-35) of their time performing social/personal tasks (e.g. sleeping, eating) and indirect care (24%; 95% CI 22-25) (e.g. reviewing notes, ordering tests). Work-related discussion comprised 15% (95% CI 13-17), and most took place at the beginning of the night. Medication-related tasks consumed a small proportion of time (4%; 95% CI 3-4) but attracted a higher level of multitasking and interruptions than most other tasks. On average, 2 h of every shift were spent at a computer and 1.3 h with patient notes. Doctors spent 72% of the night-shift alone, multitasked 6.4% of the time and were interrupted, on average, once every 46 min. CONCLUSIONS: This study provides new data about junior doctors' work at night. Relative to doctors during the day, greater proportions of time were devoted to social/personal tasks (including sleep) and indirect care, but a similar proportion to direct care. Multitasking and interruptions were minimal. Computer activities were an integral part of work. Handovers were observed at the beginning but not the completion of the night shift, which may have implications for patient safety.

Discussion of "Attitude of physicians towards automatic alerting in computerized physician order entry systems".

With these comments on the paper "Attitude of Physicians Towards Automatic Alerting in Computerized Physician Order Entry Systems", written by Martin Jung and co-authors, with Dr. Elske Ammenwerth as senior author [1], the journal wants to stimulate a broad discussion on computerized physician order entry systems. An international group of experts have been invited by the editor of Methods to comment on this paper. Each of the invited commentaries forms one section of this paper.

Audit and feedback of antibiotic use: utilising electronic prescription data.

BACKGROUND: There is now little doubt that improving antimicrobial use is necessary for the containment of resistance. OBJECTIVE: To determine whether providing individualised feedback to doctors about their recent compliance with the hospital's antibiotic policy improves compliance with the policy. METHODS: This study was conducted at a teaching hospital in Sydney, Australia. Computerised alerts integrated into the electronic prescribing system (ePS) inform prescribers of the local antibiotic policy. We utilised prescribing data extracted from the ePS for 'audit and feedback'. Thirty-six prescribers were sent feedback letters via email. We also interviewed doctors who had received letters to explore their views of the feedback and the policy in general. RESULTS: There was no significant change in compliance with the policy following implementation of the feedback intervention (0% compliant vs 11.9%; p = 0.07). Several problems with the policy and the approval process were identified by researchers during auditing and by prescribers during interviews. Some problems identified made it difficult to accurately assess compliance and for doctors to comply with the policy. CONCLUSION: Our intervention did not result in improved compliance with the antibiotic policy but revealed practical problems with the policy and approval process that had not been identified prior to the trial. Greater support for the policy by senior doctors and the assignment of more clearly defined roles and responsibilities associated with antibiotic use and approval may result in improved compliance. Harnessing the full potential of technology would streamline the antimicrobial approval process and allow more efficient and reliable monitoring of antibiotic use and compliance. Many of the problems we identified are generic issues of importance to all organisations seeking to integrate antimicrobial stewardship into ePS.