Article Text

Download PDFPDF

International Forum on Quality and Safety in Health Care, March 2009, Berlin, Germany

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


1E. N. de Vries, 2S. M. Smorenburg, 3W. S. Schlack, 1D. J. Gouma, 1M. A. Boermeester. 1Department of Surgery, Academic Medical Center, Amsterdam, The Netherlands; 2Department of Patient Safety, Academic Medical Center, Amsterdam, The Netherlands; 3Department of Anaesthesiology, Academic Medical Center, Amsterdam, The Netherlands

Background: A recent systematic review showed that over 50% of in-hospital adverse events are related to a surgical procedure. In a tertiary referral centre, a surgical safety checklist monitoring critical safety risks and information transfers during the entire surgical pathway was developed. As the first part of the step-by-step implementation of this so-called SURPASS checklist, a time out (TO) procedure was implemented. The TO was to be performed in the operating room before start of induction. The operating room assistant was to register all items of the TO, as well as intercepted incidents, in the electronic registration system already in use in the operating room.

Methods: The TO was implemented in all surgical specialties. Monthly feedback was provided to all involved personnel. All electronically self-registered intercepted incidents were extracted and analysed monthly. In addition, a sample of procedures was observed by an independent researcher to validate the data obtained by self-registration.

Results: TO compliance was high: overall performance rose from 76% in the first month to 89% in the final month. The TO intercepted many incidents: during 8923 performed TOs, 976 intercepted incidents (10.9%) were registered, among which were 21 wrong-side incidents. During the validation study, a sample of 250 procedures was observed. Of 25 observed incidents, only 9 were registered in the electronic system, implying the data from this system represent an underestimation of the number of incidents.

Conclusion: A TO procedure was successfully implemented into daily practice. Compliance was high and the TO intercepted many near-incidents, including 5 wrong-side incidents. While a time out procedure is unquestionably valuable, it is limited in scope. Not checking the essentials until inside the operating room may lead to compromised safety and/or postponement of surgery. Any intervention that aims to increase surgical patient safety should cover the entire surgical patient pathway.

Abstract 01

Number of intercepted incidents per 1000 performed time out procedures.


J. Mainz, H. Jorgensen, S. Riemann. Department of Psychiatry Region North Denmark, Odense, Denmark

The Danish National Indicator Project (DNIP) is a quality development programme based on 12 evidence based clinical indicators and adherent standards related to diagnostics, treatments and care. The indicators have been implemented in all clinical units and departments in Denmark treating patients with schizophrenia.

In 2005 clinical departments in Psychiatry in Region North achieved results significant lower than the rest of Denmark. At clinical and management level it was decided to give DNIP high priority.

In 2007/2008 “all or none” results and results related to individual indicators indicate the table has turned. The Psychiatry in Region North Denmark performs significantly above the national average and in fact celebrated results as the national best.

Valuable lessons have been learned regarding how significant quality improvements can ensured over a relative short period of time through a focused approach. Certain activities and institutional prerequisites are in our opinion essential:

  • Strong management focus on the project.

  • A continuous supply of reliable feedback (data) to support positive developments.

  • Systematically and regularly structured audit processes by audit groups consisting of clinicians at national, regional, and local levels in order to support continuous improvement of data quality, a higher degree of standard fulfilment and the implementation of improvements.

  • Development and implementation of tools (eg checklists) to support local standard fulfilment.

Though remarkable results have been achieved, there is still potential for improvement on many levels. Certain standards are still not met and as improvements are made standards will generally be raised over time.

Furthermore, the Psychiatry in Region North Denmark is working on a more fundamental level in developing relevant outcome measures making it possible to conclude more specifically on the effects of treatment and thus hopefully giving DNIP and our quality work even more legitimacy and focus.


1H. S. Chung, 1J. H. Cho, 2S. H. Hong, 2K. O. Lee, 2J. S. Lee, 2M. S. Song, 3Y. H. Cho. 1Department of Emergency Medicine, Yonsei University Gangnam Severance Hospital, Seoul, South Korea; 2Department of Nursing, Yonsei University Gangnam Severance Hospital, Seoul, South Korea; 3Department of Quality Improvement, Yonsei University Gangnam Severance Hospital, Seoul, South Korea

Background: Nurses could make errors in recognising specific medication promptly and accurately in critical patients with multiple continuous infusion medications. Emergency nurses are in a vulnerable situation to make medication errors due to high stress of workload. A system was needed to differentiate the medications continuously being infused in a single patient promptly and accurately. This work was done in an emergency department (ED) involving twenty-one nurses of a 700-bed tertiary teaching hospital in Seoul, Korea.

Assessment of Problems: This was a prospective single blinded simulation study. Three situations with six continuous infusion medications and one crystalloid fluid on a single patient were simulated. First, finding and stopping infusion of potassium mixed fluid. Second, finding and stopping heparin mixed fluid. And third, finding and loading 200 cc of crystalloid fluid. Promptness and accuracy were assessed. The colours were pre-coded by the nurse manager. After 1 week of education, the same situation was simulated with colour-coded fluids. The same nurses were assessed for promptness and accuracy for recognition of medication fluids.

Results of Assessment/Measurement: Significant improvement was noticed in promptness for all three situations before and after colour-coding: K+ mix (27.29 sec vs 7.14 sec); heparin mix (17.9 sec vs 6.31 sec); crystalloid (17.42 sec vs 6.8 sec). There were 2 inaccurate recognition of fluids for both K+ mix and heparin mix fluids.

Lessons and Messages: Being able to recognise the multiple medications being infused in a single patient can be a critical action. ED is a very busy and sometimes chaotic environment. In order to attenuate or minimise medical errors, a practical system, like the colour-coding system might be helpful. But this study was only done in a simulated patient. Further study is mandatory to see if the system improved for the operational performance of nurses.


1S. Engelmann, 2B. Köppl, 1R. Grün. 1Alice-Salomon-Hochschule, University of Applied Sciences, Berlin, Germany; 2Sana Gesundheitszentrum GmbH, Berlin, Germany

Background: Quality management (QM) in medical practices will become mandatory in Germany from 2009. We investigated the implementation of a QM system in 72 general and specialist practices in Berlin, Germany.

Assessment of Problems: While practices have instituted a range of single quality improvements in recent years, participation in systematic QM is still a challenge.

Strategy for Quality Improvement: The provider committed to implementing “Quality and Development in Practices” (QEP), a manual-based QM system that has been specifically designed for practices. The approach is based on 63 targets in the areas of patient care, patient rights and safety, practice management, and continuous quality improvement. Each practice team received intensive training and facilitated support to undertake the self-assessment and to develop its own quality manual.

Measurement of Improvement: Staff survey evaluating the training process, the quality changes in response to the project, the acceptance and use of the manual, and the willingness to engage in further QM activities.

Results of Assessment: The training events and workshops were rated as highly relevant and informative as only 20% of participants had previous knowledge of QM systems in practices. Regular practice meetings were reported by 84% of the respondents and greater transparency of management processes by 50%. Only a minority attributed better patient care and improved team coherence to the project. The practice manual was considered helpful by 48% of the respondents but only 29% had used it for reference. Willingness to engage in future QM activities was more prevalent in nurses and support staff than in doctors.

Lessons and Messages: While the implementation of QEP showed demonstrable quality improvements, there is still scepticism towards formal QM among practice members. Appropriate incentives, ongoing training, and a greater involvement of non-medical staff are important determinants in creating a culture of sustained quality improvement.


S. J. White, C. Richmond, J. Kim, S. Lowcock. Australian Commission on Safety and Quality in Health Care, Bruce, Australian Capital Territory, Australia

Clinical handover was identified as problematic in an emergency department, however previous attempts at improvement had been ineffective. External researchers suggested implementing an externally developed standardised operating protocol, however this was opposed to by staff. The decision was made to teach staff a simple ethnographic observation framework to enable them to systematise observations and understand them as evidence.

An external researcher assisted three staff members in learning ethnographic tools. After several observation sessions, a group data analysis session was held to identify common problems and develop solutions. Four handover solutions were suggested. These solutions were relevant to the local context and staff had ownership of the solutions.

This project sits in the gap between research and quality improvement. It relies on expert research guidance to frame the self-reflection of clinicians. The methodology allowed staff to view and make judgements on their local practice. By utilising this space, clinicians are empowered with the tools to effectively use qualitative research techniques to develop and implement change.

The effect of using ethnography to empower clinicians was measured by engagement of staff involved in the project and general staff enthusiasm for the staff-developed changes. The poor uptake of handover solutions was solved by allowing clinicians to develop their own solutions. The staff took ownership of the project during the first phase observations. They championed their solutions with other staff members, including senior clinicians.

This research demonstrates that handover can be successfully improved at local levels by the development of local solutions to local problems through facilitated staff research using ethnographic and reflexive techniques. By allowing staff to research handover and then develop solutions, not only do they have ownership over the solutions, but they also use their expert knowledge of the environment and culture to ensure that the solution is appropriate for the context.


K. Walters. Basingstoke and North Hampshire NHS Foundation Trust, Hampshire, UK

Background: Basingstoke and North Hampshire NHS Foundation Trust provides acute hospital services to a population of 300 000, in North Hampshire. The Trust Board was concerned that there was a lack of robust and systematic data on patient safety to provide assurance and allow the identification of areas for improvement.

Strategy for Change: To address this issue, a meeting of stakeholders was held (Directors, Senior Clinicians and Managers). A variety of patient safety metrics were proposed with outcome measures identified where possible. The metrics chosen were considered “essential” rather than “interesting” to minimise data overload. A total of 20 metrics were identified, including hospital standardised mortality ratio (HSMR), raw weekly mortality data, re-admission rates, infection rates, adverse incident rates, patient movements and patient falls. Once the metrics were identified, definitions, data sources and proposed methods of data collection were agreed. The presentation format of the metrics was a dashboard based on a traffic light scoring system to give a quick overview. More detailed statistical process control charts were provided as an appendix to the dashboard, allowing monitoring of trends.

Effect of Change: The introduction of metrics has raised the profile of patient safety at Board-level. The Board question the data and have lively discussions around concerns that the data have raised. The Board also has an increased confidence and ability to make decisions based on the data, rather than relying on anecdotal evidence. To date, the metrics have prompted several reviews, for example, a mortality review of patients with fractured neck of femur, and a review of surgical site infection in patients undergoing caesarean section. In addition, each Division of the Trust is in the process of developing their own set of metrics, relevant to the services they provide.


M. Seddon. Middlemore Hospital, Counties Manukau District Health Board, Manukau, New Zealand

The Physiologically Unstable Patient (PUP) programme was designed to recognise physiological deterioration in general ward patients, and to intervene early to prevent further deterioration. We knew from international studies that patients often displayed signs of deterioration for many hours before effective action was taken. The hospital had already instituted a Medical Emergency Team (MET), but case reports showed that this team was usually called very late in the piece when the patient was in extremis or arresting.

Three barriers were identified: (1) lack of understanding of the importance of vital sign recording (2) the recording chart for vital signs did not assist staff to identify clinical deterioration and (3) nurses did not feel confident in escalating care, or feared rebuke.

Multifaceted campaign:

  1. Design of a new vital sign chart with separate rows for each vital sign and colour-coding to highlight when a vital sign was out of the normal range;

  2. An additive early warning score (EWS) – the PUP score – developed for our patient population;

  3. A clinical escalation plan for raised PUP scores as part of vital sign chart;

  4. A nurse-led Rapid Response Team (RRT) for moderately raised PUP scores;

  5. A six-week education programme for each ward.

PUP implemented in all wards over 15 months.

The completeness of vital sign recording improved (60% pre to 90% after). The response to raised PUP scores was disappointing with only 30% of such scores being acted on. The MET callout rate showed special cause variation with an increase in callouts (from a median of 25 calls/month to 60, with 120 calls in the last month).

An EWS and nurse-led RRT can be implemented but attention must be paid to the team culture in the hospital and implementation must address the barriers that prevent staff calling for help.


I. Mitchell, H. McKay, C. VanLeuvan, S. Mamootil, C. McCutcheon, P. Lamberth, B. Avard. ACT Health, Canberra, Australia

Unexpected deaths and unplanned admissions to the intensive care unit (ICU) are often due to the failure of recognising and appropriately managing deteriorating patients. The Early Recognition of the Deteriorating Patient Program aimed to improve the recognition of deteriorating vital signs and the timeliness of appropriate medical review.

Following a literature review, three interventions were developed including a track and trigger system, a new observation chart and a locally developed COMPASS education package, which emphasised physiological reasons for measuring vital signs and a structure for communication. These interventions were initially implemented on four pilot wards then rolled out to the remaining clinical areas following the success of the pilot study.

Data were collected for three months pre and post intervention and included documentation of vital signs, the time to medical review following a physiological deterioration, unplanned admissions to ICU and hospital outcome.

The respiratory rate documentation improved from 64% to 93%. Communication relating to clinical deterioration was specifically studied in 52 patients (25 and 27 patients pre and post rollout respectively). Pre roll out, there was failure to communicate in 79% of patients but reduced 11% patient post rollout. The time for a medical review of patients was reduced from 48 minutes to 12 minutes. Medical emergency team reviews increased from 85 to 100 and intensive care transfers increased from 66 to 70 (p = 0.755). Hospital mortality reduced from 2.05% to 1.95% (p = 0.71)

Improving vital sign documentation through simple tools and appropriate education, and utilising a track and trigger system with a consistent structure for communication allows timely and appropriate medical review to facilitate the improvement in patient care.


C. R. Davis. North Bristol NHS Trust, Bristol, UK

Background: Methicillin-resistant Staphylococcus aureus (MRSA) is a global healthcare issue with medical and socioeconomic consequences for patients, healthcare professionals and hospital trusts. The main route of transmission is via direct contact. Compliance with alcohol gel is therefore essential to minimise the risk of nosocomial MRSA infections. This audit aimed to quantify and improve compliance with alcohol gel application at the entrance to a surgical ward.

Assessment of Problems: For six months, baseline compliance rates with applying alcohol-gel at a ward entrance was assessed. A discretely positioned camera was used to collect data. Gold standard compliance was set at 100% for all healthcare professionals, patients and visitors making non-emergency visits to the ward.

Results of Assessment/Measurement: Mean compliance rate of alcohol-gel application in the initial 6 months was 24% (Range 0–35%). Doctors had the lowest compliance rates whiles visitors had the highest compliance rates.

Strategies for Quality Improvement/Change: The intervention consisted of a strip of bright red tape positioned along the length of the corridor approaching the ward and pointing to the alcohol gel dispensers.

Lessons and Messages: After the intervention was in situ, mean compliance rates significantly increased to 62% (p<0.0001). Compliance of doctors and visitors improved the most after the intervention (p<0.01), whilst nurses’ compliance at 75% exceeded all groups after the intervention. There were two cases of MRSA bacteraemia in the initial six months and none in the six months after the intervention. The cost of the intervention in this audit is less than £1, resulting in improved infection-control practice and an associated eradication of MRSA. Socioeconomic costs for patients, healthcare professionals and trusts from MRSA bacteraemia far exceed this negligible investment.

Abstract 09

Alcohol gel compliance at ward entrance.


M. Heenan. St. Joseph’s Healthcare Hamilton, Hamilton, Ontario, Canada

Brief Outline of Problem: St. Joseph’s Healthcare Hamilton (Canada) developed a Medical Quality Scorecard as its quality framework to engage physicians in patient safety and quality.

Strategy for Change: While the Medical Quality Scorecard is not an exact duplicate of the balance scorecard, its fundamentals were followed during development including:

  1. Designed Medical Staff vision statement on using data for quality improvement (QI)

  2. Framed scorecard quadrants in Physician-friendly language

  3. Consulted physicians on Indicator development to create ownership

  4. Scorecard standard agenda item at Medical Staff Meetings

  5. Action QI projects using Institute for Healthcare Improvement’s Plan-Do-Study-Act (IHI-PDSA) Model on key safety issues

Quadrants are named Patient Safety and Quality; Appropriateness of Care; Resource Utilisation and Flow; and, Patient Access and Wait Times (see figure).

Effects of Changes: Scorecard assisted in improving mortality, medical documentation, infection control, ER wait times and bed management. Chiefs no longer see themselves as program representatives, but as partners engaged in medical quality.

Key Success Factors:

  1. Supportive Chief of Staff

  2. Performance Improvement Consultant for Physicians

  3. Frame issues using IHI-PDSA Model for Improvement

  4. “Paper” report created awareness and branding

  5. Chiefs compensation includes quality requirement

Message for Others: Language indicating the need to “engage” physicians in quality and patient safety can on its own deter physicians from participating in organisational wide agendas. It wrongly assumes physicians are not engaged to begin with. Those who become physicians do so to help improve patients quality of life and to protect them from harm. It’s about alignment, integration and communication – not engagement. The scorecard provides a common language that breaks down traditional silos and instils team work.

Abstract 10

SJHH Medical Quality Scorecard – quadrants and sample indicators.


B. Hoffmann, V. Mueller, D. Domschke, M. Beyer, F. M. Gerlach. Institute for General Practice, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany

Background: Patient safety culture refers to the pattern of attitudes and values surrounding the safety of patients in health care organisations. A method for a health care team to assess its safety culture, with the ultimate goal of improving it, does not yet exist in Germany. The Manchester Patient Safety Framework (MaPSaF) was primarily developed for use in general practices in the UK. Practitioners in England have found it to be a useful tool in terms of face validity and acceptance. This study aimed to adapt MaPSaF for use in the German healthcare system.

Methods: MaPSaF was translated independently by two native German speakers and both translations were consented by the project team. Face validity and acceptance of the initial version of Frankfurt Patient Safety Matrix (FraTrix) was evaluated by focus groups consisting of general practitioners and practice assistants. After readaptation of the content and structure of FraTrix, the instrument was field tested in general practices. At the end off the testing phase, field testing was evaluated by means of self-administered questionnaires and focus groups with participants.

Results: FraTrix – like MaPSaF – describes five different levels of safety culture and nine dimensions of practice. Compared to the original, the content and language of FraTrix and the setting of a team session had to be slightly modified. The adapted instrument facilitated the reflection on and discussion of the specific safety culture in every practice. Of the 19 practice teams, 15 were able to decide on an action plan in order to improve their safety culture in the future.

Lessons Learnt: FraTrix is a feasible method to initiate and facilitate discussions on patient safety culture and found acceptance among general practice teams. However, a thorough introduction of the subject (patient safety culture) and the instrument is essential if FraTrix is to work successfully.


M. Persson, E. Holm. Neonatal Unit, Sachs’ Children’s Hospital, Södersjukhuset, Stockholm, Sweden

Healthcare of today is at constant change and it’s hard for the individual to stay updated without an organisation for knowledge updates. We solved this by introducing short instruction films (Quality assurance and Learning through Instruction Films (KLIF)) and web based learning (Tool for Interactive Learning and Daily Assistance (TILDA)).

We are now one step closer to become a “High Reliability Organisation” (HRO).

In health care, there is a strong tradition of spreading knowledge verbally. However, there are some risks associated with this. Weakness lays in those situations where no one really knows how to do. This creates insecurity among staff. PM’s and guidelines rarely contain practical information. There’s also an organisational problem when staff relies on old knowledgebase. This is not consistent with high patient safety and the way of a HRO. A survey was conducted among staff to find situations where they felt insecure or lacked knowledge. We visited an ICU that had developed KLIF and TILDA to solve similar problems.

KLIF (Quality assurance and Learning through Instruction Films): The films are based on various health care routines. They increase staff knowledge and sense of confidence while carrying out routines. Involvement of staff in filmmaking (actors and reviewers) is significant for support of method and staffs confidence while working with equipment. The whole film process (research, reviewing, staff discussions) lead to more quality assurance.

TILDA (Tool for Interactive Learning and Daily Assistance): Here we keep PM’s, nursing documents and manuals on medical equipment. TILDA is divided into different qualification certificates. The employee can log in and access the evidence based material, then they answer a number of randomised questions. Staff self-govern their education and acquire new knowledge.

70 percent of staff felt more confident in their work after the introduction of KLIF and TILDA. Small mistakes decreased and the routines were carried out in a more satisfactory way. We became more of a HRO.


I. Hansen. Jonkoping County Council, Höglandssjukhuset, Eksjö, Sweden

Problem: Transformation problems of a concise patient report between staff are well known. Poor reports are harmful, reduce patient safety and elevate stress levels. Incidents lead to adverse event reports and elevated need of root cause analysis.

Strategy: The implementation of situation, background, assessment, recommendation (SBAR) started in Hoglandet Health Care Area (Jonkoping County, Sweden). The tool was demonstrated at a regional chief meeting. Information series started which included the SBAR tool addressed towards groups of doctors, nursing personal and the managing board. Four primary care districts and eight hospital departments learned about SBAR; doctors and nurses had priority. Demonstration of the tool for doctors gave very positive reactions facilitating the implementation process. Several departments would like to start using the tool at once but we need a check of quality first. One internal medicine ward tested one version during May 2008 and a pocket card in August. Next we made simulated reporting between different professionals for correct use of SBAR. With the use of interviews, the chief of an internal medicine ward measured demands and specific terms of assessment.

Improvement/Changes: The nurse staff felt very good using the SBAR, especially younger members. This was surprising compared with experience from other improvement work. From the interviews: “I feel safe when I report because I know that the risk for missing important information decreases,” “I feel safe when I contact the doctor on duty because I am well informed”, “I report present data and what is relevant in a structured fashion”, “my own information about the patient increased”.

Conclusion: Lessons learnt so far are positive. SBAR is easy to use and the concept seems easy to “sell” to doctors, nurses and managing leaders. The pocket card format is convenient to users.


R. J. Mead, A. Tobin, M. Atkins, N. Bannerjen, M. Duku, D. Pearl, A. Chauhan. Portsmouth Hospitals NHS Trust, NHS Education South Central, Portsmouth, UK

Wessex NHS and Portsmouth Hospitals Trust are engaging and facilitating junior doctors, with junior doctor led improvement projects.

A review of elective orthopaedic patients in Surgical High Care (SHC) at Royal Hospital Haslar (RHH), together with informal surveys of medical, surgical and nursing teams identified recognition of medical disease, diagnostic speed, and management as contributory factors in medical cause admissions. Our challenge was improving medical knowledge in a novel fashion because EWT compliant rotas drastically reduce teaching opportunities.

Junior surgical staff were surveyed for solutions and an interactive e-mail based set of cases and multiple-choice questionnaires developed. These allowed medical learning, case based discussions, and practice towards surgical examinations.

The e-cases were sent 4 weekly, and juniors asked to return answers. Answers and discussion were provided a week later. A twice daily handover allowed reinforcement of teaching and individual discussions.

The medical and surgical teams were re-surveyed after 4 cases. SHC admission cases were audited.

All 14 juniors confirmed receipt, thirteen reporting no formal teaching on these topics prior to the e-cases, and all 14 were positive about the cases. The five RMOs all reported improvement over 4 months.

The percentage of medical cases admitted fell by 2% over the same period, all cause referral rose 33%.

Flexible interactive e-teaching can complement medical education, it can be delivered at times suiting the individual, the material can be relevant to routine practice, the uptake of cases can be measured, and compliance with education can be supported by material that helps in higher professional exams. This is likely to improve the quality of health outcome but now needs formal testing.

Cross faculty and speciality co-operation and teaching has the potential to significantly improve doctor knowledge and routine medical care in surgical patients.

E-based learning can help timetabling difficulties associated with modern hospital practice.


L. L. Liang. Kaiser Permanente, USA

Kaiser Permanente, a US based health plan and integrated delivery system implemented an ambulatory electronic medical record system and personal health record in four years for 8.6 million patients. Completion of the electronic medical records for their 30+ hospitals will be complete by year end 2009. They reported a dramatic shift in modalities to communicate and receive ambulatory healthcare in one of their eight regions. Total office visits decreased 26% with a decline of 25.3% in general medicine and 21.5% in specialty consultant visits while overall clinical contacts increased 8.3% due to additional secure email contacts and scheduled phone visits. Clinical outcomes and patient satisfaction were stable during this timeframe. Roughly 3 million patients actively use the personal health record features of secure email to clinicians; medication refills; and online access to past medical visits, immunisations, laboratory test results, patient support programs, and general health information. Approximately 500 000 secure email messages are sent to Kaiser Permanente physicians each month with responses in 24–48 hours. These results identify a significant opportunity to redesign healthcare more efficiently and effectively. Patients and physicians alike report greater familiarity with each other and relevant medical issues. However, creating value by implementing electronic medical records requires fundamental redesign of daily work processes in health care delivery reinforced by appropriate financial investments and incentives.


J. Stoves, J. Connolly, A. Grange, R. Roberts, J. Wright. Bradford Institute for Health Research, Bradford, UK

Problem and Context: The primary care electronic health record (EHR) is not accessible to secondary care staff, consequently paper letters and telephone calls are the principal means of communication across the interface between primary and secondary care. A preliminary audit of paper referrals to the local renal service showed that many were inappropriate and/or lacked important clinical information. The Chronic Kidney Disease Electronic Advisory Service (CKDEAS) was therefore established to improve communication between hospital kidney specialists and primary care physicians (GPs).

Strategy for Change: CKDEAS was piloted in a single GP practice and then evaluated in 15 early adopter practices. Participating GPs attended education events and received paper and e-guidance.

Measurement of Improvement: Quantitative Analyses included the number, appropriateness and quality of GP referrals and the timeliness of response by the kidney specialist. A qualitative evaluation of GP satisfaction with CKDEAS was also performed.

Effects of Changes: E-consultations lasted 15 minutes and the response time was 5 days (52 days for paper). Urinalysis data were available in 77% of patients (31% for paper) and there were more creatinine data (8 vs 1.9 for paper). CKDEAS was easy to use, quick and convenient, provided helpful and succinct advice and avoided outpatient referrals. GPs reported more confidence in managing CKD patients in the community.

Lessons Learnt: E-consultation permits a detailed and efficient review of the primary care EHR. Patients requiring hospital clinic review are promptly identified and others benefit from timely specialist advice. CKDEAS will be made available to all practices, and e-communication using the primary care EHR is to be introduced throughout the renal department. The hospital IT infrastructure will be expanded to support commissioning of e-consultation for renal services, and there is interest in developing e-consultation for other chronic disease management services in the region.


A. Jewkes, V. Varadarajan, A. Watson. Department of General Surgery, Manchester Royal Infirmary University Hospital, Manchester, UK

Background: Concerns were raised over record keeping in our University General Surgery department. It was commonly difficult to elucidate issues such as the timing of decisions and patient identifiers on record sheets.

Assessment of Problems: We formulated departmental guidelines based on Royal College of Physicians recommendations. Minimum standards required the presence of three patient identifiers (patient name/hospital number/ward location) and seven key entries regarding the previous team review (date/time of review, senior present, name/signature/job title/contact number of recording doctor). Patient records were randomised and the last surgical entry was manually reviewed.

Results: Particular deficits were noted in identification numbers (noted in 38% of records), ward location (0%), time of entry (20%), and the name, job title and contact number of the documenting practitioner (44%, 42% and 60%).

Strategies for Change: A local three-month departmental campaign was launched, entitled “Document”. Junior grades were contacted with new guidelines. Posters were displayed on wards, raising awareness amongst all professionals. Senior surgeons were personally asked to lead by example. Re-auditing post-campaign, statistically significant improvements were seen in patient name (recorded in 100% of notes vs 78%), identification number (70% vs 38), ward location (38% vs 0%), time of entry (67.5% vs 20%), and name (87.5 vs 44%)/contact number (80% vs 60%, p = 0.03) of the documenting practitioner (all p<0.01, Fisher’s exact test, unless stated). Feedback showed staff could now identify/contact relevant individuals, and handover was more efficient, with additional patient identifiers in place.

Lessons and Messages: With minimal resources and outlay, a significant improvement in the standard of record keeping was created in a short time period. In an increasingly evolving health system, focus on small scale local changes and communication are often lost. However, statistically significant improvements are clearly achievable at this level with minimal expenditure.


K. H. Seng. Institute of Mental Health/Woodbridge Hospital, Singapore

Background: In IMH, there are 3 psychogeriatric long-stay wards with 140 patients (18 patients on depot and oral antipsychotics) and 24 adult long-stay wards with 958 patients (555 patients on depot and oral antipsychotics). The patients, not discharged due to social reasons, were stable and mainly reviewed by medical officers rotated 6 monthly. Medications were largely not readjusted. Majority do not need polypharmacy due to aging, supervised environment and chronicity of illness. Polypharmacy led to side effects, increased cost and administration time.

Assessment of Problems and Results: A team including psychiatrists, medical officers, nurses, pharmacist and patient embarked on a project: To reduce antipsychotic polypharmacy1 in all eligible2 psychogeriatric long-stay patients within 6 months, without having any relapses.

  1. On 2 or more antipsychotics (including depot).

  2. In remission of psychotic symptoms for at least 6 months.

Through brainstorming and survey of patients, doctors and nurses, problems identified included lack of guidelines, lack of continuity of care by rotating medical officers who may lack experience and doctors’ and nurses’ fear of relapse.

Strategies for Improvement: Simple guidelines and a dose reduction table were developed and training provided for medical officers. Plan-do-study-act methodology was utilised.

Effects of Changes: All 8 eligible patients had antipsychotics reduced, 4 had depot taken off and mean antipsychotic dosage was reduced by more than half. No patient relapsed, 3 had improved blood pressure, 3 had extrapyramidal side effects reduced and 2 became more alert. Medical Officers were more confident and staff and patients expressed satisfaction. There were drug cost savings and time saved for nursing care.

Lessons: With guidance, junior doctors were empowered to reduce polypharmacy safely. Constant education and evidence-based results are essential to change people’s mind-set. The project led to spread to adult psychiatry long-stay wards and will be applied to future long-stay patients.


Wolff, S. Taylor, A. McGrath. Clinical Risk Management Unit, Wimmera Health Care Group, Wimmera, Australia

Brief Outline of Context: Stroke inpatients in a rural base hospital in Victoria, Australia between 1999 and 2007.

Brief Outline of Problem: Process indicators are sensitive measures of quality of patient care. Some patients do not receive all processes of care that would benefit them. An “all or none measurement” of compliance with all key process indicators provides greater precision in measuring quality of care.

Assessment of Problem and Analysis of Its Causes: Checklists and reminders for process indicators were incorporated into clinical pathways for stroke patients. Could improvements be sustained? What proportion of patients received all the key processes of care?

Strategy for Change: A multidisciplinary clinical team adapted best practice evidence for stroke for local conditions. Processes of care were provided to clinicians as clinical pathways incorporating checklists and reminders, integrated into the patients’ medical record and completed by clinicians as they provide care. Results were regularly fed back to all clinical staff.

Measurement of Improvement: The proportion of patients with stroke who received individual and all key processes of care before and after the introduction of the pathway.

Effects of Changes:

Abstract 19

Timeline of proportion of patients with stroke who received key interventions before and after introduction of the clinical pathway.

Lessons Learnt: Key success factors were:

  • Each process of care was provided as checklists and reminders incorporated into the patient medical record and completed by clinical staff providing care.

  • Clinical pathways were multidisciplinary, developed improving communication and team work and provided ownership.

  • Funding was provided for the program’s coordination.

  • Medical staff were involved early in pathway development and before implementation.

  • There was an established clinical risk management culture.

  • Clinical and executive champions steered the pathway program through hospital clinical and administrative systems.

Message for Others: Clinical pathways incorporating checklists and reminders improve and sustain quality of patient care. The “all or none measurement” approach provides added precision in measuring improvement.


R. S. Zimmerman, E. M. Oswald. Mayo Clinic, Phoenix, Arizona

Problem: Root cause analyses (RCA) at Mayo Clinic Arizona (MCA) identified opportunities for improvement in provider communication. Unfortunately no process assessed specific gaps in, or effectiveness of providers’ communication. An immediate solution was also necessary when facing communication failures.

Assessment: Intimidation, failure to support staff, and a lack of clear communication were recurring contributing factors. To reduce patient harm events and mortality, we coined “meta-communication” to represent our concept of “communicating about our communication” just as meta-data is data about data.

Strategy: A “Plus One” pilot program established the principal tenet: “It is always appropriate and desirable for any caregiver to communicate directly and openly with any level of the medical/surgical team to assure a timely assessment of and plan for a patient’s condition”. These performance guidelines and cultural expectations were endorsed by institutional leadership. “Plus One” reports are generated whenever a caregiver needs to “work” or “add someone” to get help. MCA’s “Plus One” program is a rapid response methodology with a feedback loop for “communication about the communication”. When applicable, “Plus One” reports are referred to residency program directors and department chairs. Analyses provided to leadership identify trends and factors giving rise to “Plus One” scenarios, such as day, time, service/specialty and individual resident/attending involvement.

Measurement: The MCA Board formalised the program in 2005 and measurement showed continued improvement with a 67% reduction in harm events to date. In its first year (’05–’06) the “Plus One” meta-communication program contributed to a significant improvement in hospital standardised mortality ratio (HSMR) (p<0.05). Since then MCA has improved its HSMR yearly, remaining significantly lower than the national average (p<0.05, fig. 1).

Messages: Empowerment and support of line staff by leadership is required to overcome communication obstacles. Communication about our communication (“meta-communication”) is essential to reduce patient harm and provide safe care.

Abstract 20

Risk-adjusted mortality rates (HSMR).


N. El Hosseiny, A. El Ansary, B. Abd ElAziz, M. El Saidy. Dar AlFouad Hospital, Giza, Egypt

Brief Outline of Context: Appropriate medication management system in hospitals is an integral part of patient safety. Dar AlFouad Hospital drug committee decided to introduce a unit pharmacists service to improve two important steps in the medication service which are drug dispensing and preparation. A team was formed in May 2008 to study and implement proactively this new service.

Brief Outline of Problem: During auditing by Clinical pharmacists:

  1. Medication dispensing error represented 67% of total medication errors during 2007

  2. Percentile of Accurate Medication Preparation 65%

Strategy for Change: DAFH Hospital decided hiring unit pharmacists for the units, available 24 hrs/day/7 days/week for drug ordering and preparing the unit doses starting May 08.

  • Proposed Flow Chart of the process

  • Proactive FMEA methodology

Abstract 21 FMEA analysis of one of the steps in the process

Effects of Changes: Data analyses after 6 months:

  • Effect of change on drug dispensing: Medication dispensing error decreased to 30% of total medication errors

  • Effect of change on drug preparation: Accurate medication preparation steps improved to 99%

Lessons Learnt: Using FMEA methodology in a team work manner has a very beneficial effect to avoid any pitfalls and reduce wastes.


A. Rubinstein, F. Rubinstein, M. Botargues, M. Barani, K. Kopitowski. Division of Family and Community Medicine, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina

Introduction: As a way to improve the quality of health care provided, the practice of paying physicians on quality targets has spread rapidly in developed countries. Little is known about quality improvement strategies that include P4P in Latin America. Furthermore, there are no reports that evaluate clinical performance of primary care providers in our country.

Objective: To report the development, design, implementation and first results of this intervention in our group.

Population and Setting: The participants are the 65 full time family practitioners (FPs) of the Division of Family and Community Medicine at Hospital Italiano.

Intervention: In 2005, the Program of Quality Improvement (PQI) started a multimodal intervention based on teamwork, continuous education, audit and feedback to improve quality of care provided. At the same time, a new scheme of P4P started to operate as a complement of the capitation reimbursement scheme. FP were organised in five groups (UDAs) that met weekly to hold on clinical discussions, update clinical topics, have periodic feedback of the routine audits, and to work on strategy development to manage identified problems in patient care. Regarding to the financial incentive topic, each UDA can earn up to 1000 points depending on the degree of accomplishment on different targets in a complex set of indicators pertaining to five different dimensions of patient care. At the end of the year, the PQI assigns to each UDA the points earned, that are then “translated” into Argentinean pesos which are paid to each participant.

Results: By the end of 2007, the UDAs achieved 60–80% of all possible points. The reward in pesos would represent the equivalent of 4% of the mean year income for direct patient care.

Abstract 22

Conclusions: Financial incentives were only one component of a broader strategy that included audit and feed-back, education, teamwork and peer pressure, which improved quality of patient care.


M. Kotagal, P. Lee, C. Habiyakare, R. Dusabe, H. M. Epino, P. Kanama, M. L. Rich, P. E. Farmer. Partners In Health/Inshuti Mu Buzima (PIH/IMB), Rwinkwavu, Rwanda

Problem: Hospitals in rural Africa, such as in Rwanda, often lack electricity, supplies, and personnel. In this setting, basic care processes, including vital sign monitoring, medication administration, and laboratory test execution, were being performed unreliably, causing delays in treatment.

Setting: This improvement project was undertaken in a 50-bed district hospital in rural Rwanda.

Design: Simple quality improvement tools, including PDSA cycles and Failure Modes and Effects Analysis, were utilised to improve system-level processes in a stepwise fashion, in conjunction with resource augmentation as necessary.

Measurement of Improvement: Three indicators (percentage of vital signs taken by 9 am, medications administered as prescribed, and laboratory tests performed and documented) were tracked daily. Data were collected from a random sample of 25 charts from six inpatient wards.

Strategy for Change: Our intervention had two components: staff education on quality improvement and routine care processes, and stepwise implementation of system interventions. Real-time performance data were reported to staff daily, with a goal of 95% performance for each indicator within two weeks. A Rwandan QI team was trained to run the hospital’s QI initiatives.

Results: Baseline and post-intervention means show that, within two weeks, all indicators reached the 95% goal. Within one month, they were consistently 100%. Doctors and nurses subjectively reported improved patient care and staff morale.

Lessons Learnt: Five lessons are highlighted. (1) Making the data visible and using them to inform subsequent interventions can promote change in resource-constrained settings. (2) While improvements can be made in advance of resource inputs, sustained change in resource-constrained settings requires additional resources. (3) Local leadership is essential for success. (4) On-the-job learning, in lieu of didactic trainings which take staff away from the hospital, allows for efficient training of quality improvement staff in under-resourced settings. (5) Early successes are essential for motivating staff buy-in.

Abstract 23 Means and p values for routine care indicators pre and post intervention


J. T. Paige, V. Kozmenko, R. Paragi Gururaja, T. Yang, S. W. Chauvin. Louisiana State University Health Sciences Center, New Orleans, Louisiana, USA

Background: The System for Teamwork Effectiveness and Patient Safety (STEPS) was developed with support from the Agency for Healthcare Research and Quality (AHRQ) to address the documented lack of a patient safety culture and recognised poor teamwork among healthcare professionals in the US, especially in high risk locales like the operating room (OR). For this presentation, we examine aspects of its impact at the largest state-run hospital in southern Louisiana.

Assessment of Problem: Due to its high visibility, the OR was targeted for the initial hospital introduction of STEPS. A validated AHRQ endorsed Hospital Survey on Patient Safety Culture (HSPSC) was used to quantify patient safety culture at baseline and post-intervention. Its change was measured at the hospital-wide and surgery department levels; statistical analysis was conducted to compare pre- and post-intervention mean scale scores.

Strategy for Change: Using a novel mobile mock operating room (MMOR) configuration, STEPS teamwork training was implemented at the point of care in the hospital OR using high-fidelity simulation. Every member of the general surgical OR teams participated in two half-day training modules. Session 1 introduced core teamwork behavioural competencies. Session 2 targeted a specific pre-operative briefing strategy to promote teamwork behaviours. Training was distributed and involved structured debriefings and deliberate practice. Staff in-service meetings, focus groups, and feedback discussions with hospital and OR leaders occurred before, during, and after training. Observations and training evaluations were also used to examine STEPS’s short-term influence on teamwork behaviours.

Lessons and Messages: Although some improvement was noted between pre- and post-HSPSC mean scale scores, few statistically significant differences existed at either the hospital or surgery department level. Some short-term behaviour changes were noted within OR teams. Cultural change is a slow process and is typically preceded by behavioural change. Organisational support is essential for successful change process.

Abstract 24


B. Tai, A. Munns, H. Ward, on behalf of the eDS (electronic Discharge Summary) Working Party

Sponsoring Institution: The Prince Charles Hospital, Brisbane, Australia

Background: In February 2007, the electronic Discharge Summary (eDS) project was implemented at The Prince Charles Hospital (TPCH), in consultation with doctors, nurses, pharmacists, administrative staff and general practitioners (GP).

Methods: The eDS is interfaced with two hospital databases:

  1. Hospital Based Computer Information System (HBCIS) records patient information including adverse drug reactions (ADR), which is uploaded automatically into eDS.

  2. Pharmacy Medication List (PML), in which pharmacists record medication information and generate medication lists for all patients on discharge. ADR is also recorded, but not uploaded into eDS.

On discharge, pharmacists reconcile discharge prescriptions, updated medication information in PML and activated electronic authorisation in PML. The doctor could retrieve authorised medication information from PML and use as the discharge medication record (DMR) in the discharge summary. However, the doctor could also complete the. DMR by entering medication information manually in eDS.

Results: Initial audit (n = 129) revealed 40% of prescriptions accurately reflected the discharge medication regimen intended. Pharmacists reconciled 98% of prescriptions, of which 95% were entered in PML accurately. 72% of 121 discharge summaries were sent with a DMR, of which 48% were accurate. ADR documentation was 84% accurate in PML, whereas 19% accurate in discharge summaries. Subsequently eDS was modified so the DMR in the discharge summary can only be pharmacist-reconciled medication information authorised in PML. Alternatively, the discharge summary is sent with a notation “medication list to be finalised”. Doctors cannot enter medication information manually in the discharge summary. Re-audit (n = 48) showed 98% accuracy in PML. 90% of 40 discharge summaries were sent with a DMR, of which 94% were accurate.

Lessons Learnt: Introduction of a forcing function ensured pharmacist-reconciled medication information was communicated to patients and GP accurately. Auditing identified problems and provided evidence for resource allocation in areas needing support.


J. Phillips, J. Wheelan, J. Wheelan, H. Parsons, J. Branter, L. Hunt. Taunton and Somerset NHS Foundation Trust, Taunton, Somerset, UK

Brief Outline of Context: Taunton Hospital has 15 operating theatres. As part of the Safer Patient Initiative the “peri-operative” team aimed to improve team culture and reduce surgical site infections over a 2-year period.

Brief Outline of Problem: Surgical site infection can be reduced by introducing a “surgical site infection” care bundle consisting of: prophylactic antibiotics 1 hour pre knife to skin, maintaining peri-operative normothermia, and achieving glycaemic control in diabetics.

Assessment of Problem: Data collection monitored the following:

  1. Patient temperature in recovery greater than 36°C.

  2. Antibiotics given within 60 minutes before skin incision.

  3. Blood glucose on admission to recovery between 4–13 mmol/l if patient diabetic.

  4. A pre-operative safety briefing.

Strategy for Change: Reliable processes were developed using rapid cycle change in one theatre area. When reliability had been achieved these were then disseminated across the theatre complexes.Structured safety briefings were introduced before the start of every operating list.

Measurement of improvement: Run charts have shown that:

  • Safety briefings are occurring in all theatres with 85% to 100% reliability.

  • Antibiotics on time with 95% reliability.

  • Normothermia in recovery achieved with 90–95% reliability.

  • Glycaemic control has been improved.

Effects of Changes: Reliable processes have been developed. Communication has improved and safety briefings are now part of the operating theatre culture.

Lessons Learnt: Data collection and feedback is essential to motivate team members and provide evidence of change. It is important to select an effective multi-disciplinary team and change that team according to the needs of the project.

Message for Others: The quality of patient care can be improved by developing reliable processes. A change in theatre culture has been possible through a multi-disciplinary team approach.


J. Halm, J. Jordan, M. Gennis, R. Schoenenberger, A. Staroszczy. UW School of Medicine and Public Health, Milwaukee Campus, Aurora Sinai Medical Center, Milwaukee, Wisconsin, USA

Brief Outline of Context: A 195-bed acute care metropolitan teaching hospital utilised Healthcare Failure Mode and Effects Analysis (HFMEA) to reduce risks related to inaccuracies from poor medication reconciliation (MR) and communication at hospital discharge.

Brief Outline of Problem: Patients are most at risk for adverse drug events during transitions of care including hospital discharge.

Assessment of Problem and Analysis of Its Causes: The HFMEA determined the causes of inaccuracies and failure to deliver discharge medication list (DML) to community healthcare providers (CHP) were: incorrect or missing CHP information; discrepancies among actual prescriptions written and dictation summaries; patients not receiving a DML; non-compliance with use of MR form; poor communication among providers regarding discharge medications.

Strategy for Change: Education on importance of accurate and complete DML information to CHP; pharmacy technicians to assist with admission MR use; prompt dictation of DML; nurse provides patient with a plastic “My Medication Folder” to hold copy of DML, and instruction to take to all healthcare visits; pilot with selected physicians with subsequent adoption by all physicians.

Measurement of Improvement: Run charts of percentage of patients with completed admission MR forms; quarterly report on percent reconciled DML and percent of discharged patients who return to the hospital or CHP with their plastic medication folder.

Effects of Changes: MR form use improved from 67% to 92%; DML was 95% reconciled with written prescriptions, a 35% improvement. The problem of inaccurate DML was significantly resolved for heart failure care management initiative. Patient care was improved with the provision of an accurate and complete DML to patients and their CHP.

Lessons Learned: An accurate admission MR form provides an accurate DML on discharge; know system capabilities for an efficient transition.

Message for Others: Interdisciplinary approach and patient involvement in process are essential for successful implementation.


W. Kirschner. Forschung Beratung + Evaluation FB + E, Berlin, Germany

Preterm birth is internationally and nationally one of the leading problems in obstetrics. In Germany actually 8.6% of pregnants are affected. Preterm birth is the leading cause of infant mortality and surviving infants have higher morbidity levels all their live. Preterm births are generating high costs in the health system.

Epidemiological research on the risk factors of preterm birth has consistently derived a set of factors, which increase the risk of preterm deliveries such as smoking, vaginal infections, stress. To emphasise of the overall 12 risk factors only the age, the social class of women and the sex of the infant are not accessible to prevention. The evidence with respect to working interventions however turned out to be relatively small.

Interventions can be uni- or multifactorial. Besides the instrumental conception the time of intervention will be different ranging from preconceptional over prenatal to antepartal programs. To summarise we developed two programmes to reduce preterm deliveries by individual health counselling and promotion: BabyCare, which was introduced in obstetric care in Germany in 2000, followed by the preconceptional programme planbaby in 2007.

Annually the rate of preterm births of participants of BabyCare is evaluated compared to the prenatal database of lower Saxony, which stands for the average rate in all 16 German countries. Controlled for age, parity, education level and multiples we are registering a stable reduction of the preterm birth rate between 25% and 27%. Up to now more than 120 000 pregnants have participated.

Preterm birth can be reduced significantly. Important is the co-operation of health insurance companies actively supporting the programme by media coverage and funding. A multifactorial approach to reduce preterm deliveries is effective and efficient. Program coverage however must be still increased.


H. M. Seidling, S. P. W. Schmitt, T. Bruckner, J. Kaltschmidt, M. G. Pruszydlo, C. Senger, T. Bertsche, I. Walter-Sack, W. E. Haefeli. Department of Internal Medicine VI, University of Heidelberg, Heidelberg, Germany

While medication errors threatening patient safety may occur at all stages of treatment, drug prescribing is particularly error-prone, most often due to prescription of excessive doses. Clinical decision support (CDS) systems haven shown to improve prescribing quality, especially when outflanking their weaknesses by presentation of clinically relevant warnings. We developed a CDS system providing immediate feedback on prescribed dosages and considering individual patient characteristics thus increasing alert specificity. The CDS system was integrated into the local electronic prescribing platform. A comprehensive software algorithm compared the prescribed daily dose with a personalised upper dose limit which was calculated considering individual patient characteristics (eg age, renal function, co-medication) as stored in the electronic patient chart. Whenever a prescription was classified as overdosed, the physician was invited to modify the dosage regimen or, when insisting on the original dosage regimen, to specify his reasons. The benefit of the CDS system was evaluated in a two-phased prospective study (phase 1: baseline, no feedback; phase 2: intervention, immediate dosage feedback) by comparing the number of excessive doses in the two phases. During phase 1, 4.5% of the prescriptions (552/12 197) were categorised as overdosed compared to 559 prescriptions initially triggering a warning in phase 2 (4.8%; similar to the 4.5% in phase 1 (p = 0.37)). In response to this alert, physicians modified the dosage regimen or switched to an alternative drug in 134 cases, decreasing the number of finally prescribed excessive doses to 425 (3.6%), thus reducing the overdose rate by 20% compared to baseline (p<0.001). Physicians’ adherence depended on patients’ age and was higher for certain drug classes (eg antineoplastic drugs). The CDS system did not induce new manifest medication errors (assessed as potential underdose and underuse), suggesting that a highly patient-specific CDS system safely improves prescribing quality.


K. Freeman, R. Byrne, G. Caldwell. Worthing and Southlands Hospitals NHS Trust, West Sussex, UK

Patient satisfaction is widely used as a marker of success; however, staff satisfaction is less commonly assessed. We describe the use of simultaneous staff and patient satisfaction surveys to facilitate improvements in service provision.

Worthing Hospital is 600-bed hospital on the south coast of England, providing acute services to around 300 000 people. Acute medical admissions were increasing, and a pilot Direct Admissions Area (DAA) was set up, with the aim of streamlining the admissions process. Staff working in DAA expressed dissatisfaction with the pilot and we sought to more formally assess these problems.

We designed staff and patient satisfaction surveys. Different problems were highlighted by each group. Staff were concerned about location and lack of space. Patients were generally more satisfied but expressed concern about lack of refreshments. The data were presented to managers and changes made. A new area was designated DAA, with more beds and equipment, and faster access to diagnostics.

The surveys were repeated one month later. The second round of questioning revealed higher levels of satisfaction amongst staff and patients. 74% of staff (previously 34%) now agreed that DAA was in a suitable location and 66% believed that the new location provided a satisfactory amount of space to examine patients (0% previously). Overall, the percentage of DAA staff rating their satisfaction levels as good or very good increased from 3% to 66%. Patient satisfaction was consistently high; 87% of patients reported an overall satisfactory experience in DAA initially and 95% after the changes were implemented.

During periods of changes in service it is important to assess satisfaction of both patients and healthcare providers. Opinions can be obtained simultaneously, rapidly and relatively simply. Patient and staff opinions do not necessarily correlate, but by failing to obtain views from both, key suggestions for improvement may not be obtained.


K. M. Maney, S. H. Chang, P. Tilakaratna, A. Wantman. Barts and The London NHS Trust, London, UK

Background: Hospital acquired infections are a significant cause of morbidity and mortality. Hand cleansing is one of the most effective preventative measures. In response to a perceived poor compliance in hand cleaning practices, we did a series of audits at The Royal London Hospital (a large teaching hospital).

Assessment of Problem: We carried out a questionnaire survey, observational audits on hand hygiene practices of anaesthetists in anaesthetic and recovery rooms, including microbiological swabs from 30 anaesthetists.

Results: Questionnaire survey: most claimed to have good hand hygiene practices. In anaesthetic room, 48% of the anaesthetist cleaned their hands before induction. Prior to intravenous cannulation and airway management 79% and 81% worn gloves respectively. In recovery room, 58% wore gloves when transferring the patient and 42% cleaned their hands before leaving recovery room. No pathogens were grown from the ID badges or mobile phones of anaesthetists; one hand imprint grew coagulase negative staphylococci.

Strategy for Change: Results of our audit were presented and the importance of hand hygiene was highlighted to all. Guidelines for good hand hygiene practices were placed throughout the theatre complex.

Measurement of Improvement: A repeat audit was done. The results were:

  • In anaesthetic room, 85% cleaned their hands prior to induction, 37% increase. 65% and 75% wore gloves for IV cannulation and airway management respectively, a decrease in compliance.

  • In recovery room, 50% brought the patients wearing gloves and 30% of them washed their hands, a decrease in compliance.

Lessons Learnt: Education and guidelines can make a positive difference but these efforts may need to be repeated regularly. Highlighting the difference between hand hygiene with hand cleaning and personnel protection with universal barrier precautions is also vital.

Message for Others: Hand cleaning plays an important role in reducing hospital-acquired infections. Frequent reminders can lead to increase in compliance.


S. Marsch, F. Tschan, N. K. Semmer, R. Zobrist, P. Hunziker, S. Hunziker. Department of Medical Intensive Care, University of Basel, Basel, Switzerland

Background: There are two obstacles to fully understand the processes leading to medical error: (1) lack of feedback as the error becomes not necessarily obvious during the further course and, (2) retrospective analysis limited by hindsight bias. A prospective analysis, starting prior to the chain of events eventually leading to a medical error, is therefore warranted to gain more insight into the mechanisms of medical errors. The aim of the study was to establish whether medical simulation is suitable to expose and investigate medical errors.

Methods and Measurements: A high-fidelity human patient simulator was used. 40 teams of three hospital physicians each were randomised to two versions of a penicillin-triggered anaphylactic shock that developed according to a pre-programmed script: in version “distraction” the patient’s complaints were suggestive of a tension pneumothorax; the tension pneumothorax however could be simply ruled out by normal breath sounds during auscultation; the version “normal” had no in-build distraction. All analysis was performed post-hoc using video-recordings obtained during simulations.

Results: 17/20 teams in the “normal” version but only 6/20 teams in the “distraction” version respectively were able to correctly diagnose and treat the anaphylactic shock (p = 0.001). Most failing teams stuck to their initial diagnosis and were unable to consider alternative hypotheses (fixation error). In many teams the interpretation of breath sounds obtained by auscultation depended on the current hypothesis of the team rather than objective findings (breath sounds were kept constant during the scenario) indicating a perception bias.

Lessons and Messages: Medical simulation allows the investigations on medical errors that for medical, ethical and practical reasons cannot be performed in real patients. Fixation error can be easily provoked and is neither considered nor detected by the physicians involved. The current mental model rather than objective findings determines the interpretation of “objective” medical information.

Abstract 32

Quality of diagnosis and treatment in teams faced with one of two versions of a simulated anaphylactic shock. The two versions significantly (p = 0.001) differed with regard to correct diagnosis and correct treatment.


M. B. Gibson, R. Verrier-Jones, C. Fuller. NESC (NHS Education South Central), Salisbury NHS Foundation Trust, Salisbury, UK

Background: We are a group of newly qualified doctors working in an acute general hospital in which we learnt about quality improvement by working in supervised action learning sets. During our first weeks, finding our way around the wards and locating specific items had been frustrating when covering unfamiliar wards. Our aim was to reduce the time spent hunting for the patient, their notes, and investigation request cards.

Assessment of Problem: Process map of steps taken when called to assess a patient.

  • Photographic journal noted considerable variation in the organisation of bed numbering, whiteboards, notes trolley labels, and the location of investigation request cards on different wards.

  • Time trials to find a patient, their notes, and necessary request cards varied three fold from best to worst (44 sec–149 sec).

  • Questionnaires reported frustration on a daily basis with regards to hunting and gathering.

Results of Assessments: We attended Sisters’ meetings and asked for three volunteer wards to trial our strategies together with a ward clerk who would champion these improvements.

Mechanism for Improvement: Layout and information given on whiteboards was standardised, notes trolleys were re-labelled, and wards received a bright yellow 24-drawer stationary cabinet holding all doctors’ forms in identical order.

Measurement of Improvement: Time trials showed a reduction in the time spent hunting. Questionnaires indicated less frustration, and nurses noted fewer interruptions from doctors asking for help to locate stationary and notes, which were now more frequently refilled in the correct place.

Lessons and Messages: Difficulty and time to implement simple changes into a large organisation.

  • Only tackle a problem you feel strongly about and believe can be changed.

  • Implement small changes at a time.

  • Keep people informed of every step so as not to offend anyone.

  • Perseverance and commitment will succeed.


B. Ho, H. Fung, L. Chung, W. L. Au, K. T. Tam, D. Lam, M. Chui, P. L. Chan, J. Liu, I. Lam, H. Lee, S. Chan, B. Wong, A. Chan, J. Cheng, S. F. Lui. New Territories East Cluster, Hong Kong

Background: In 2005, with the aim of improving medication safety, the New Territories East Cluster of Hong Kong observed the “3 Checks 5 Rights” practice amongst different nurses and noticed the inconsistency.

Assessment of Problem and Results: A survey was conducted in which about seventy nurses were invited to describe “3 Checks 5 Rights”. It was noticed that they gave ten different versions on first check, as well as eleven versions on the second and third checks. In reviewing the literature, there was also no expert consensus on the details of “3 Checks”.

Strategy for Quality Improvement/Change: An initiative was undertaken with the aims of: (a) standardising the checking process and (b) improving the outcome by reducing errors. Several principles were adopted during the redesign: (a) refer to practical situations and workflow; (b) emphasise on “doing the things right for the first time” and reduce repetition; and (c) develop a mental picture to facilitate easy memorisation of the process. The revised “3 Checks 5 Rights” was put into effect in April 2006. The nursing drug administration incidence reduced by 32% and 38% in the first and second year respectively after implementation. In the staff survey, more than 90% respondents agreed the revised procedure provided clearer information, was more practical and adequate in ensuring safe drug administration. Very high compliance was noted in the audit.

Lessons and Message: Though “3 Checks 5 Rights” is a legendary nursing practice, it should be updated where necessary in order to be effective in achieving safe drug administration. Standardisation has also demonstrated its effectiveness in preventing error. However, the credibility cannot be entirely given to the revised process as other factors such as training or increased staff awareness may have also contributed to the improvement.


G. Miller, B. Dean Franklin, A. Jacklin. Imperial College Healthcare NHS Trust, London, UK

Background: In UK hospitals, traditional ward pharmacy services are provided by pharmacists undertaking daily visits to their allocated ward(s), where they monitor prescriptions, mainly on a retrospective basis. Pharmacists discuss medication related issues and make recommendations to medical staff where necessary; this is typically termed an intervention. As many interventions are made to resolve a prescribing error or to the improve quality of care, it is important that interventions are made as soon as possible after prescribing, or preferably, at the point of prescribing. With the traditional service, there can be a delay between the prescription being written and a pharmacist’s intervention taking place.

Strategy for Change: We have been increasing the number of specialist pharmacists who routinely attend consultant led ward rounds.

Measurement of Improvement:Design: Prospective, non-randomised, controlled study. Setting: Five inpatient medical wards at two teaching hospitals. Study aim: Compare the number, nature and clinical importance of interventions made by pharmacists attending consultant led ward rounds in addition to providing a clinical ward pharmacy service (study group), with those made by pharmacists providing a traditional clinical ward pharmacy service alone (control group).

Result of Assessment: A mean of 1.73 physician accepted interventions were made per patient for the study group, compared to 0.89 for the control (Mann Whitney U, p<0.001). There was no difference between groups in the nature or clinical importance of the interventions. Each consultant led ward round lasted on average 115 minutes, during which one physician accepted intervention was made every eight minutes, compared to one every 63 minutes during a ward pharmacist visit which had a mean duration of 68 minutes.

Lessons and Messages: Pharmacists attending consultant led ward rounds in addition to undertaking a ward pharmacist visit make significantly more interventions per patient, thereby reducing preventable medication errors and optimising treatment.


V. Dhungana, B. Sthapit. Kathmandu Medical College Teaching Hospital, Kathmandu, Nepal

Background: There are huge differences between quality in healthcare in developing and developed nations which affects patients’ satisfaction and compliance to healthcare they receive. A study was done with the objective to identify the gaps in quality in healthcare in Nepal and explore measures to improve the situation. It was done in a medical OPD of a tertiary-level-hospital of Nepal where ten doctors attend around 250 patients each day.

Assessment of Problem: A cross-sectional-survey was done among OPD-patients regarding the patients’ expectations and satisfaction from the services.

Results of Assessment/Measurement: Survey revealed that most doctors (90% in one way or the other) were not dealing with patients as they expected and 36% were dissatisfaction with the services. This was a serious issue that could affect patient compliance and outcomes.

Strategies for Quality Improvement/Change: Results of the study were discussed and strategies proposed in the hospital. An effective written policy with patient centeredness and engagement was developed. Another study was done to measure the feasibility of the new policy, in which a group of doctors were oriented and patients’ survey done after two months.

Lessons and Messages: Patient satisfaction and compliance in follow up and treatment significantly increased with new strategy. Doctors welcomed the changes with ease. 8.4% of patients however said that they were confused due to treatment choices given to them; reason observed was illiteracy among that particular group. The only challenge was the decreased numbers of about 20 patients per day the OPD could now accommodate due to more time spent per person. It was solved by reducing clerical works by doctors with the help of assistants in the OPD. Gaps in quality in healthcare were identified and strategies proposed and adopted were feasible; giving an example for other developing nations too. Increasing patient engagement makes patients aware, satisfied and more responsible towards their health and can improve outcome of treatment.


J. E. Klopotowska, R. A. J. Kuiper, H. J. M. van Kan, A. C. J. M. de Pont, M. G. W. Dijkgraaf, S. M. Smorenburg, L. Lie-a-Huen, M. B. Vroom. Academic Medical Centre, Amsterdam, The Netherlands

Brief Outline of Problem: Patients admitted to intensive care unit (ICU) are at high risk for prescribing errors. The effect of active participation of clinical pharmacists in an ICU team on prescribing errors is unknown in a Dutch hospital setting.

Assessment of Problem and Analysis of Its Causes: All consecutive patients admitted to the adult ICU of Academic Medical Centre in Amsterdam from 1 September 2005 to 1 July 2006 where included. The study was divided into 3 phases: phase 1 (baseline), phase 2 (introduction of intervention) and phase 3 (stabilisation of intervention). After of an ICU training period of 4 weeks, the clinical pharmacist reviewed 3 times a week all new drug prescriptions, formulated recommendations for improvement of the drug regiments prescribed and discussed this recommendation during the daily patient review with the attending ICU physicians. These activities are referred to as the intervention.

Measurement of Improvement: We analysed and compared the outcome measures between the 3 study phases. Only the recommendations accepted by the attending ICU physicians were counted as prescribing errors.

Effects of Changes: The incidence of prescribing errors found: 190/1000 during phase 1, 78/1000 during phase 2 (difference of 112/1000, 95% CI 80/1000 to 144/1000, p<0.001), and 45/1000 during phase 3 (difference of 33/1000, 95% CI 11/1000 to 55/1000, p = 0.004). The pharmacist made 659 recommendations during the whole study period, of which 439 (67%) were accepted by the physicians, which resulted in changing the wrong drug prescriptions according to the advice of the pharmacist. The rate of acceptance was 60.4% during phase 1, 74.1% during phase 2 and 73.8% during phase 3.

Lessons Learnt: Reviewing the medication and participation of a clinical pharmacist in a multidisciplinary ICU team in the Netherlands was associated with significant reduction of prescribing errors. The recommendations were well accepted by the ICU physicians.


G. J. Shortland, M. Smithies, G. Ellis. University Hospital of Wales, Heath Park, Cardiff, UK

Introduction: Participation in the Safer Patients Initiative Phase II (SPI II) has provided an opportunity to guide a pilot project for a medical rapid response team (MRRT) supporting three adult medical wards.

Methods: The service comprises of five nurses based within the Directorate of Integrated Medicine. They can be called to the three medical wards by staff who have concerns about a patient’s condition based on clinical concern and/or a Modified Early Warning Score (MEWS) in normal working hours. The results of a six month pilot are reported.

Results: Patients observation frequency at the beginning and end of the pilot improved, with respiratory rate (31% vs 98%), AVPU (45% vs 100%) and temperature (87.5 vs 100%). A MEWS score was recorded in 99.1% of patients. Where a change in frequency of observations was recommended following a patient contact this was implemented in 89% of instances. Attendance of the team resulted in 70% having some form of management change (24% had oxygen implemented, 25% were given I.V. fluids and 27% a change in drug therapy). Eight per cent of patients seen (total seen 224) had a palliation plan initiated, 12.5% had a DNAR order initiated, 25% had a referral for specialist assessment. The MRRT was able to document that the change in the care plan was implemented in 92% of cases within 4 hours.

Discussion: The establishment of the MRRT within a pilot area is able to positively influence taking and recording of observations, the use of a physiological trigger tool and ensure appropriate changes in required monitoring of patients observations and implement changes in management.


1J. Bacou, 2P. Garel, 1C. Bruneau, 1R. Amalberti. 1National Authority for Health, Saint-Denis La Plaine Cedex, France; 2European Hospital and Healthcare Federation (HOPE), Brussels, Belgium

The European Union Network for Patient Safety (EUNetPaS) is a network involving the 27 member states of the European Union, funded and supported by the European Commission within its 2007 Public Health Programme. This project is coordinated by HAS (French National Authority for Health). Its purpose is to establish an umbrella network of all 27 EU Member States and EU stakeholders to encourage and enhance their collaboration in patient safety (culture, reporting and learning systems, medication safety and education) through the establishment of national platforms involving national stakeholders.

EUNetPaS will promote coherence at EU level through recommendations and proposition of common tools.

  • Culture measurement tool

  • Guidelines for education

  • Library of methods for reporting and learning systems implementation

  • Rapid response mechanism for sharing high priority patient safety issues or solutions between all member states

  • Medication safety recommendation which will be tested in health care organisations in 10 member states. The objectives are to identify good practices to reduce medication errors in hospitals which could be implemented in other national environments. More than 60 good practices have been identified in the first semester 2008 and seven of them have been selected. The field implementation will run for a 9 month period beginning April 1st 2009.


D. Anoff, L. Guerra, M. Morgan, R. Gross. Moffitt Cancer Center, Tampa, Florida, USA

The Institute for Healthcare Improvement’s (IHI) 100 000 Lives Campaign was designed to save 100 000 lives over 18 months and projected that 60 000 lives could be saved with Rapid Response Teams (RRT) alone. Thus, RRT at Moffitt Cancer Center was implemented 02/20/07.

The goal was a reduction in morbidity/mortality and improving outcomes by reducing the time elapsed before medical intervention was delivered following signs of distress, thereby preventing deterioration to cardiopulmonary arrest. Target population did not include Code Blues.

Data on codes/month, survival of event, survival to discharge and location were compared before and after implementation of RRT.

Clinicians designated RRT members, created policy and documentation. RRT members participated on an on-call schedule 24/7.

IHI data previously demonstrated that 66% to 75% of patients showed clinical signs of deterioration 6–8 hours prior to Code Blues. The goal was to see a reduction in Code Blues and in mortality 6 months post-implementation.

Individuals could initiate RRT calls if concerned for a patient’s immediate safety and if the patient had a pulse and respirations. Calls were documented on the RRT-SBAR (Situation, Background of Event, Assessment, and Recommendation). Call #/month, reason for call, DNR status and average call length were recorded.

The year prior to RRT’s implementation was compare to the first year post implementation:

  • Total Code Blues decreased from 78 pre-RRT to 48 post-implementation (38% decrease).

  • Survival of event increased from 83 to 87.5%.

  • Survival to discharge increased from 45 to 80%.

RRT reduced response time for patients in distress thereby improving outcomes.

Abstract 40

Number of code blues Jan 2006–Feb 2008.