Search

Healthcare worldwide is faced with a crisis of patient safety: every day, everywhere, patients are injured during the course of their care. Notwithstanding occasional successes in relation to specific harms, safety as a system characteristic has remained elusive. We propose that one neglected reason why the safety problem has proved so stubborn is that healthcare suffers from a pathology known in the public administration literature as the problem of many hands. It is a problem that arises in contexts where multiple actors – organizations, individuals, groups – each contribute to effects seen at system level, but it remains difficult to hold any single actor responsible for these effects. Efforts by individual actors, including local quality improvement projects, may have the paradoxical effect of undermining system safety. Many challenges cannot be resolved by individual organisations, since they require whole-sector coordination and action. We call for recognition of the problem of many hands and for attention to be given to how it might most optimally be addressed in a healthcare context.

Preventable harm is a major cause of preventable death worldwide. In late 1999, the Institute of Medicine (IOM) released To Err is Human ,1 a report that riveted the world's attention to between 44 000 and 98 000 patient deaths annually in the USA from medical errors. Progress towards reducing these harms has proven difficult because healthcare lacks robust mechanisms to routinely measure the problem and estimates of the magnitude vary widely. It is hard to gauge safety when healthcare uses multiple different measures for the same harm and provides limited investment in measurement, implementation and applied sciences. Central line-associated bloodstream infection (CLABSI) provides a notable exception and case study for learning. Over the past 15 years, through the combined and coordinated efforts of many, these infections have been reduced over 80% in intensive care units (ICU), decreasing patient mortality.2 In this essay, we reflect on the journey in preventing these infections and explore how this success can inform and accelerate efforts to reduce other types of preventable harm. While this paper is a synthesis of past work, what is novel is providing the historical profile of CLABSI, comparing infection rates before and 15 years after the IOM report and offering new insights into what led to the substantial reductions in infections. The journey began in the 1970s when the Centers for Disease Control and Prevention (CDC) began collecting data on ICU CLABSIs. Over the next several decades, the CDC published data on national benchmarks for CLABSI, investigated bloodstream infection outbreaks and published the first Guideline for the Prevention of Intravascular Catheter-Related Infections in 1991. These early efforts brought little change in ICU CLABSI rates throughout the 1990s, with clinicians and policy makers believing infections were inevitable. Over the last decade, ICU CLABSI rates have dropped throughout the USA (table 1) and …

Pressure injuries are problematic to health systems throughout the world, significantly harming over 7 million patients and adding extraordinary costs. The USA, for instance, experiences more than 2.5 million pressure injury cases per year which cause over 60 000 deaths—that is more than car accident fatalities in the USA—and cost the health system at least $9–$11 billion.1 The UK is no less affected by 700 000 cases per year that result in 27 000 deaths and cost the National Health Service (NHS) an estimated £1.4–£2.1 billion.2 Despite there being no reason to expect major differences in the presentation of pressure injuries between the two countries, the UK appears mainly to have a problem with chronic pressure ulcers, whereas the USA appears to have an increasing burden of acute pressure ulcers.3 4 The UK NHS reports chronic wounds accounting for 78% of the NHS spending on wound treatments.5 On the other hand, the acute issue in the USA may be connected to the Centers for Medicare and Medicaid Services (CMS) reimbursement policy, which has cut payments since 2008 for hospitalised patients who acquired pressure injuries as well as penalised facilities with poorest hospital-acquired condition rates.6 7 This CMS policy also focuses financial responsibility of the pressure injury on the facility that identifies the wound, which means that facilities trying to correct other’s mistakes are often blamed. Like most quality measures derived from billing data that have a financial penalty, variability in pressure injury rates is a combination of changes in reporting and how clinicians care for patients by implementing a pressure injury prevention protocol. While we do not have empirical evidence for the relative proportions that these two domains contribute to pressure injury rates, our experience suggests that health systems’ initial response to CMS policy has been improvements in coding and tracking pressure injuries. …

One of the key recommendations of the Institute of Medicine9s (IOM) report, To Err is Human , 15 years ago was for greater attention to incident reporting in healthcare, analogous to the role it has played in aviation and other high-risk industries. With the passage of time and maturation of the patient safety field, we conducted semistructured interviews with 11 international patient safety experts with knowledge of the US healthcare and meeting at least one of the following criteria: (1) involved in the development of the IOM9s recommendations, (2) responsible for the design and/or implementation of national or regional incident reporting systems, (3) conducted research on patient safety/incident reporting at a national level. Five key challenges emerged to explain why incident reporting has not reached its potential: poor processing of incident reports (triaging, analysis, recommendations), inadequate engagement of doctors, insufficient subsequent visible action, inadequate funding and institutional support of incident reporting systems and inadequate usage of evolving health information technology. Leading patient safety experts acknowledge the current challenges of incident reports. The future of incident reporting lies in targeted incident reporting, effective triaging and robust analysis of the incident reports and meaningful engagement of doctors. Incident reporting must be coupled with visible, sustainable action and linkage of incident reports to the electronic health record. If the healthcare industry wants to learn from its mistakes, miss or near miss events, it will need to take incident reporting as seriously as the health budget.

Improvement (defined broadly as purposive efforts to secure positive change) has become an increasingly important activity and field of inquiry within healthcare. This article offers an overview of possible methods for the study of improvement interventions. The choice of available designs is wide, but debates continue about how far improvement efforts can be simultaneously practical (aimed at producing change) and scientific (aimed at producing new knowledge), and whether the distinction between the practical and the scientific is a real and useful one. Quality improvement projects tend to be applied and, in some senses, self-evaluating. They are not necessarily directed at generating new knowledge, but reports of such projects if well conducted and cautious in their inferences may be of considerable value. They can be distinguished heuristically from research studies, which are motivated by and set out explicitly to test a hypothesis, or otherwise generate new knowledge, and from formal evaluations of improvement projects. We discuss variants of trial designs, quasi-experimental designs, systematic reviews, programme evaluations, process evaluations, qualitative studies, and economic evaluations. We note that designs that are better suited to the evaluation of clearly defined and static interventions may be adopted without giving sufficient attention to the challenges associated with the dynamic nature of improvement interventions and their interactions with contextual factors. Reconciling pragmatism and research rigour is highly desirable in the study of improvement. Trade-offs need to be made wisely, taking into account the objectives involved and inferences to be made.

Objective Behavioural marker systems are advocated as a method for providing accurate assessments, directing feedback and determining the impact of teamwork improvement initiatives. The present article reports on the state of quality surrounding their use in healthcare and discusses the implications of these findings for future research, development and application. In doing so, this article provides a practical resource where marker systems can be selected and evaluated based on their strengths and limitations. Methods Four research questions framed this review: what are the attributes of behavioural marker systems? What evidence of reliability and validity exists? What skills and expertise are required for their use? How have they been applied to investigate the relationship between teamwork and other constructs? Results Behavioural markers systems are generally designed for specific work domains or tasks. They often cover similar content with inconsistent terminology, which complicates the comparison of research findings across clinical domains. Although several approaches were used to establish the reliability and validity of marker systems, the marker system literature, as a whole, requires more robust reliability and validity evidence. The impact of rater training on rater proficiency was mixed, but evidence suggests that improvements can be made over time. Conclusions A consensus of definitions for teamwork constructs must be reached to ensure that the meaning behind behavioural measurement is understood across disciplines, work domains and task types. Future development efforts should focus on the cost effectiveness and feasibility of measurement tools including time spent training raters. Further, standards for the testing and reporting of psychometric evidence must be established. Last, a library of tools should be generated around whether the instrument measures general or domain-specific behaviours.

Peer-reviewed publication plays important roles in disseminating research findings, developing generalisable knowledge and garnering recognition for authors and institutions. Nonetheless, many bemoan the whole manuscript writing process, intimidated by the arbitrary and somewhat opaque conventions.This paper offers practical advice about organising and writing a manuscript reporting quality improvement or patient safety research for submission to a peer-reviewed journal.Each section of the paper discusses a specific manuscript component-from title, abstract and each section of the manuscript body, through to reference list and tables and figures-explaining key principles, offering content organisation tips and providing an example of how this section may read. The paper also offers a checklist of common mistakes to avoid in a manuscript.

Background We sought to characterise the frequency, health outcomes and economic consequences of diagnostic errors in the USA through analysis of closed, paid malpractice claims. Methods We analysed diagnosis-related claims from the National Practitioner Data Bank (1986–2010). We describe error type, outcome severity and payments (in 2011 US dollars), comparing diagnostic errors to other malpractice allegation groups and inpatient to outpatient within diagnostic errors. Results We analysed 350 706 paid claims. Diagnostic errors (n=100 249) were the leading type (28.6%) and accounted for the highest proportion of total payments (35.2%). The most frequent outcomes were death, significant permanent injury, major permanent injury and minor permanent injury. Diagnostic errors more often resulted in death than other allegation groups (40.9% vs 23.9%, p

Context: Misdiagnoses may be an underappreciated cause of preventable morbidity and mortality in the intensive care unit (ICU). Their prevalence, nature, and impact remain largely unknown. Objectives: To determine whether potentially fatal ICU misdiagnoses would be more common than in the general inpatient population (w5%), and would involve more infections or vascular events. Data sources: Systematic review of studies identified by electronic (MEDLINE, etc.) and manual searches (references in eligible articles) without language restriction (1966 through 2011). Study selection and data abstraction: Observational studies examining autopsy-confirmed diagnostic errors in the adult ICU were included. Studies analysing misdiagnosis of one specific disease were excluded. Study results (autopsy rate, misdiagnosis prevalence, Goldman error class, diseases misdiagnosed) were abstracted and descriptive statistics calculated. We modelled the prevalence of Class I (potentially lethal) misdiagnoses as a non-linear function of the autopsy rate. Results: Of 276 screened abstracts, 31 studies describing 5863 autopsies (median rate 43%) were analysed. The prevalence of misdiagnoses ranged from 5.5%e100% with 28% of autopsies reporting at least one misdiagnosis and 8% identifying a Class I diagnostic error. The projected prevalence of Class I misdiagnoses for a hypothetical autopsy rate of 100% was 6.3% (95% CI 4.0% to 7.5%). Vascular events and infections were the leading lethal misdiagnoses (41% each). The most common individual Class I misdiagnoses were PE, MI, pneumonia, and aspergillosis. Conclusions: Our data suggest that as many as 40500 adult patients in an ICU in USA may die with an ICU misdiagnoses annually. Despite this, diagnostic errors receive relatively little attention and research funding. Future studies should seek to prospectively measure the prevalence and impact of diagnostic errors and potential strategies to reduce them.

As the importance of patient safety has been broadly acknowledged, various improvement programmes have been developed. Many of the programmes with proven efficacy have been disseminated internationally. However, some of those attempts may encounter unexpected cross-cultural obstacles and may fail to harvest the expected success. Each country has different cultural background that has shaped the behavior of the constituents for centuries. It is crucial to take into account these cultural differences in effectively disseminating these programmes. As an organ transplantation requires tissue-compatibility between the donor and the recipient, there needs to be compatibility between the country where the program was originally developed and the nation implementing the program. Though no detailed guidelines exist to predict success, small-scale pilot tests can help evaluate whether a safety programme will work in a new cultural environment. Furthermore, a pilot programme helps reveal the source of potential conflict, so we can modify the original programme accordingly to better suit the culture to which it is to be applied. In addition to programme protocols, information about the cultural context of the disseminated programme should be conveyed during dissemination. Original programme designers should work closely with partnering countries to ensure that modifications do not jeopardise the original intention of the programme. By following this approach, we might limit barriers originating from cultural differences and increase the likelihood of success in cross-cultural dissemination.

ObjectiveTo describe the important distinctions between traditional biomedical and high reliability organisation (HRO) approaches in advancing healthcare delivery. To further describe how these two concepts can be integrated in a learning health system to improve care for patients, providers’ experience and healthcare organisations’ effectiveness.Methods and analysisA description of how HRO and traditional biomedical approaches address problems in healthcare is provided, along with a description of their key characteristics. This is followed by a discussion on how to combine these concepts synergistically to the benefit of all stakeholders in a learning health system.ResultsHealth systems are challenged to become learning health systems but the path toward this goal is not well described, particularly on how to integrate existing approaches.ConclusionHRO and traditional biomedical approaches are important strategies to improving healthcare. However, they fundamentally address the challenges in healthcare differently. It is important to find ways to integrate these concepts to achieve maximum benefit in a learning health system.

Progress toward improving patient safety has been slow despite engagement of the health care community in improvement efforts. A potential reason for this sluggish pace is the inadequate integration of human factors and ergonomics principles and methods in these efforts. Patient safety problems are complex and rarely caused by one factor or component of a work system. Thus, health care would benefit from human factors and ergonomics evaluations to systematically identify the problems, prioritize the right ones, and develop effective and practical solutions. This paper gives an overview of the discipline of human factors and ergonomics and describes its role in improving patient safety. We provide examples of how human factors and ergonomics principles and methods have improved both care processes and patient outcomes. We provide five major recommendations to better integrate human factors and ergonomics in patient safety improvement efforts: build capacity among health care workers to understand human factors and ergonomics, create market forces that demand the integration of human factors and ergonomics design principles into medical technologies, increase the number of human factors and ergonomic practitioners in health care organizations, expand investments in improvement efforts informed by human factors and ergonomics, and support interdisciplinary research to improve patient safety. In conclusion, human factors and ergonomics must play a more prominent role in health care if we want to increase the pace in improving patient safety.

Background Differences in contexts (eg, policies, healthcare organisation characteristics) may explain variations in the effects of patient safety practice (PSP) implementations. However, knowledge of which contextual features are important determinants of PSP effectiveness is limited and consensus is lacking on a taxonomy of which contexts matter. Methods Iterative, formal discussions were held with a 22-member technical expert panel composed of experts or leaders in patient safety, healthcare systems, and methods. First, potentially important contextual features were identified, focusing on five PSPs. Then, two surveys were conducted to determine the context likely to influence PSP implementations. Results The panel reached a consensus on a taxonomy of four broad domains of contextual features important for PSP implementations: safety culture, teamwork and leadership involvement; structural organisational characteristics (eg, size, organisational complexity or financial status); external factors (eg, financial or performance incentives or PSP regulations); and availability of implementation and management tools (eg, training organisational incentives). Panelists also tended to rate specific patient safety culture, teamwork and leadership contexts as high priority for assessing their effects on PSP implementations, but tended to rate specific organisational characteristic contexts as high priority only for use in PSP evaluations. Panelists appeared split on whether specific external factors and implementation/management tools were important for assessment or only description. Conclusion This work can guide research commissioners and evaluators on the contextual features of PSP implementations that are important to report or evaluate. It represents a first step towards developing guidelines on contexts in PSP implementation evaluations. However, the science of context measurement needs maturing.

Background Implementing bundles of best practices has been shown to provide patients with recommended care and reduce medical errors. Rhode Island9s (RI) hospital leaders, quality organisations and insurers discussed the results of a quality improvement initiative in Michigan, the Keystone project, and explored the possibility of replicating these results statewide in RI. Design Hospital executives and intensive care unit (ICU) staff, RI9s quality organisations, RI Quality Institute, Quality Partners of RI, and Hospital Association of RI and consultants from Johns Hopkins University, worked together to implement evidence-based interventions and change safety culture in RI9s ICUs. Objectives The authors describe the RI ICU Collaborative, funded by insurers and hospitals, and report on statewide central line-associated bloodstream infection (CLABSI) and ventilator-associated pneumonia (VAP) rates between 1 January 2006 and 30 June 2008. Subjects All adult ICU patients in RI. Measures CLABSI and VAP rates. Results 100% of 23 ICUs in 11 hospitals participated in the RI ICU Collaborative. The statewide mean CLABSI rate decreased 74% from 3.73 (median 1.95) infections per 1000 catheter days to 0.97 (median 0) in quarter (Q) 2 (March–June) 2008 (p=0.0032). The VAP rate fell 15% from 3.44 (median 0.58) to 2.92 VAPs (median 0) per 1000 ventilator days in Q2, 2008. Conclusion The RI ICU Collaborative, a statewide quality improvement initiative, served as the platform by which multifaceted interventions were associated with reductions in CLABSI and VAP rates, and an increase in the use of evidence-based interventions. Completing Phase II, the RI ICU Collaborative continues to sustain these statewide reductions.

Objective: To develop a practical set of measures for routine monitoring, performance feedback, and improvement in the quality of palliative care in the intensive care unit (ICU). Design: Use of an interdisciplinary iterative process to create a prototype “bundle” of indicators within previously established domains of ICU palliative care quality; operationalization of indicators as specified measures; and pilot implementation to evaluate feasibility and baseline ICU performance. Setting: The national Transformation of the Intensive Care Unit program developed in the United States by VHA Inc. Patients: Critically ill patients in ICUs for 1, >3, and >5 days. Measurements and main results: Palliative care processes including identification of patient preferences and decision making surrogates, communication between clinicians and patients/families, social and spiritual support, and pain assessment and management, as documented in medical records. Application is triggered by specified lengths of ICU stay. Pilot testing in 19 ICUs (review of >100 patients’ records) documented feasibility, while revealing opportunities for quality improvement in clinician-patient/family communication and other key components of ICU palliative care. Conclusions: The new bundle of measures is a prototype for routine measurement of the quality of palliative care in the ICU. Further investigation is needed to confirm associations between measured processes and outcomes of importance to patients and families, as well as other aspects of validity.

Objective: To assess the cost-effectiveness of a multifaceted quality improvement programme focused on reducing central line-associated bloodstream infections in intensive care units.\ud \ud Design: Cost-effectiveness analysis using a decision tree model to compare programme to non-programme intensive care units.\ud \ud Setting: USA.\ud \ud Population: Adult patients in the intensive care unit.\ud \ud Costs: Economic costs of the programme and of central line-associated bloodstream infections were estimated from the perspective of the hospital and presented in 2013 US dollars.\ud \ud Main outcome measures: Central line-associated bloodstream infections prevented, deaths averted due to central line-associated bloodstream infections prevented, and incremental cost-effectiveness ratios. Probabilistic sensitivity analysis was performed.\ud \ud Results: Compared with current practice, the programme is strongly dominant and reduces bloodstream infections and deaths at no additional cost. The probabilistic sensitivity analysis showed that there was an almost 80% probability that the programme reduces bloodstream infections and the infections’ economic costs to hospitals. The opportunity cost of a bloodstream infection to a hospital was the most important model parameter in these analyses.\ud \ud Conclusions: This multifaceted quality improvement programme, as it is currently implemented by hospitals on an increasingly large scale in the USA, likely reduces the economic costs of central line-associated bloodstream infections for US hospitals. Awareness among hospitals about the programme's benefits should enhance implementation. The programme's implementation has the potential to substantially reduce morbidity, mortality and economic costs associated with central line-associated bloodstream infections.

Medicine has an inefficient knowledge market. Patients often receive half of the recommended therapies and it takes over 15 years for evidence to be incorporated into routine practice. The IOM report on comparative effectiveness research highlighted the need to produce evidence that is useful to decision makers.1 In medicine evidence is usually summarized in practice guidelines, which are needed and often elegant and scholarly literature summaries, developed by professional organisations or government agencies are used to shape clinical practice standards. The purpose of these guidelines is to help ensure that patients receive recommended therapies, yet often they are neither used nor as useful as the developers would like.2 As Schnoor et al illustrate ( see page 498 ), it is often difficult to demonstrate that guidelines improve patient care even with an implementation strategy in place.3 This may be because of clinician barriers (they do not know about or agree with the evidence), guideline barriers (the guideline may be ambiguous or not associated with outcomes), system barriers (the supplies needed to comply with the guideline may not be readily available) or implementation barriers (the way the guideline was introduced engendered resistance).4 We propose that guidelines should not be viewed as a final product to guide clinical practice but rather as an input for a more succinct tool—a checklist. Checklists are not Harry Potter's wand and when used alone will probably be no more effective than guidelines.5 Yet when used as part of a conceptual model to identify local barriers to compliance, to measure and feedback performance, and to change the local work system to ensure patients receive the checklist items, it can lead to significant improvement.6 7 In this essay, we discuss some …

Problem Venous thromboembolism (VTE) is a common cause of potentially preventable mortality, morbidity, and increased medical costs. Risk-appropriate prophylaxis can prevent most VTE events, but only a small fraction of patients at risk receive this treatment. Design Prospective quality improvement programme. Setting Johns Hopkins Hospital, Baltimore, Maryland, USA. Strategies for change A multidisciplinary team established a VTE Prevention Collaborative in 2005. The collaborative applied the four step TRIP (translating research into practice) model to develop and implement a mandatory clinical decision support tool for VTE risk stratification and risk-appropriate VTE prophylaxis for all hospitalised adult patients. Initially, paper based VTE order sets were implemented, which were then converted into 16 specialty-specific, mandatory, computerised, clinical decision support modules. Key measures for improvement VTE risk stratification within 24 hours of hospital admission and provision of risk-appropriate, evidence based VTE prophylaxis. Effects of change The VTE team was able to increase VTE risk assessment and ordering of risk-appropriate prophylaxis with paper based order sets to a limited extent, but achieved higher compliance with a computerised clinical decision support tool and the data feedback which it enabled. Risk-appropriate VTE prophylaxis increased from 26% to 80% for surgical patients and from 25% to 92% for medical patients in 2011. Lessons learnt A computerised clinical decision support tool can increase VTE risk stratification and risk-appropriate VTE prophylaxis among hospitalised adult patients admitted to a large urban academic medical centre. It is important to ensure the tool is part of the clinician’s normal workflow, is mandatory (computerised forcing function), and offers the requisite modules needed for every clinical specialty.

The need to improve the quality of care is well recognised. Yet accomplishing this is complicated, messy, and uncertain, requiring that researchers tackle technical (science) and adaptive (emotional, social, cultural, and political) challenges.1 Tension exists between those who say “just do something” to improve quality and those who say “science should be the guide.”2 The two linked studies (doi:10.1136/bmj.d195; doi:10.1136/bmj.d199) suggest that more science is needed. Benning and colleagues evaluated a large patient safety programme (the Safer Patients Initiative; SPI) in the United Kingdom, led by the Institute for Healthcare Improvement.3 4 The Health Foundation initiated and supported the initiative and, laudably, an independent evaluation, grounded in theory and conducted by experts in epidemiology, biostatistics, medical sociology, health services research, and clinical medicine. They performed a quantitative and qualitative evaluation at organisational and patient levels. The evaluation included five substudies that looked at whether the interventions worked and why. In addition to using a rigorous research design, the authors conducted a state of the art analysis, which included using different approaches to evaluate changes over time in treatment and comparison hospitals. This evaluation will serve as a model for the field. It required, however, an interdisciplinary team of experts and appropriate research funding, both of which are rare​rare. The study’s findings are partly encouraging and partly worrying. On the encouraging side, the study provides convincing evidence that safety and quality improved in NHS hospitals in England over the study period (about 18 months). This should provide comfort to UK citizens, the NHS, and parliament. Patients are less likely to be harmed from the care they receive. The NHS should try to understand why these improvements occurred and how they can be strengthened and replicated broadly across the UK. For those who hoped SPI would transform care, the findings are disconcerting. The authors found that the initiative had no discernible additional effect on patient safety; care improved to the same extent in both treatment and comparison hospitals, highlighting the need for robust evaluation with concurrent controls. It is, of course, difficult to measure the impact of patient safety interventions, especially diffuse interventions like SPI. The initiative might possibly have provided benefits that were not measured or may emerge over time. It is also difficult in these types of large scale evaluations to find an appropriate comparison group. In areas where intervention hospitals were performing well at baseline, it would be difficult to show improvement. The study should be a wakeup call to those implementing patient safety programmes. Too many patients in the UK, and the rest of the world, continue to experience preventable harm. The quality improvement field needs to embrace science, favour evidence over anecdote, and move beyond using only one generic framework for improvement (the plan, do, study, act cycle).5 Different types of patient safety challenges exist, such as translating evidence into practice, improving teamwork and organisational culture, identifying and mitigating hazards, and reducing diagnostic errors. Each type of problem should be informed by specific theories, methods, and measures.6 Although well intentioned, it is not surprising that the SPI had less of an impact than the investigators anticipated. It was not grounded in a theory of organisational change.7 It asked hospitals to implement 43 interventions, when most hospitals would find it difficult to implement three. Clinicians thought that many of the interventions were supported by weak evidence and that some measures were not valid. The initiative was largely top down, with limited input from local clinicians. Moreover, it did not target areas where teams performed poorly (in many of the areas, teams were performing nearly flawlessly before the initiative). The interventions and measures were not sufficiently pilot tested, and quality control over the quality improvement data collected by the local teams was virtually non-existent. Clinicians who push back against patient safety interventions are often viewed as “knaves.”8 This study suggests that some of that resistance may be warranted. Some interventions focused on areas that were not problematic and used evidence and measures that doctors did not always perceive as valid, potentially souring clinicians’ attitudes towards efforts to improve patient safety. We need clinicians to lead patient safety efforts. For this to happen, they must believe that interventions and measures are based on science and that their patients will benefit. Yet when interventions deal with both technical and adaptive challenges, broad scale improvement in patient safety is possible. Several patient safety interventions have shown significant improvements in patient outcomes by having clinicians and researchers collaborate when developing and pilot testing the programme. Such innovations include centralised collection of performance measures and evidence summaries by the researchers, and local innovation of programme implementation by the clinicians.9 10 11 12 These studies provide three important lessons. Firstly, patient safety studies require robust design and evaluation.13 Funding agencies need to support the development and implementation of patient safety programmes that include rigorous evaluation. These programmes should be grounded in change theory, include evidence based interventions, valid measures, and data quality control. Although theory and interventions evolve over time, patient safety programmes should be developed in collaboration with clinicians and be pilot tested, and measures should be validated before broad implementation. Secondly, care in the UK is improving; we should understand how and why. Thirdly, quality improvement efforts must improve, embracing rather than running from science. The science for quality improvement differs from basic and clinical research. It requires input from clinicians, health service researchers, social scientists, and human factors and systems engineers; in addition, it uses change theory, mixed methods, and robust evaluation. These studies provide a model.

Objective To evaluate whether implementation of the Michigan Keystone ICU project, a comprehensive statewide quality improvement initiative focused on reduction of infections, was associated with reductions in hospital mortality and length of stay for adults aged 65 or more admitted to intensive care units. Design Retrospective comparative study, using data from Medicare claims. Setting Michigan and Midwest region, United States. Population The study period (October 2001 to December 2006) spanned two years before the project was initiated to 22 months after its implementation. The study sample included hospital admissions for patients treated in 95 study hospitals in Michigan (238 937 total admissions) compared with 364 hospitals in the surrounding Midwest region (1 091 547 total admissions). Main outcome measures Hospital mortality and length of hospital stay. Results The overall trajectory of mortality outcomes differed significantly between the two groups upon implementation of the project (Wald test χ 2 =8.73, P=0.033). Reductions in mortality were significantly greater for the study group than for the comparison group 1-12 months (odds ratio 0.83, 95% confidence interval 0.79 to 0.87 v 0.88, 0.85 to 0.90, P=0.041) and 13-22 months (0.76, 0.72 to 0.81 v 0.84, 0.81 to 0.86, P=0.007) after implementation of the project. The overall trajectory of length of stay did not differ significantly between the groups upon implementation of the project (Wald test χ 2 =2.05, P=0.560). Group differences in adjusted length of stay compared with baseline did not reach significance during implementation of the project (−0.45 days, 95% confidence interval −0.62 to −0.28 v −0.35, −0.52 to −0.19) or during post-implementation months 1-12 (−0.59, −0.80 to −0.37 v −0.42, −0.59 to −0.25) and 13-22 (−0.67, −0.91 to −0.43 v −0.54, −0.72 to −0.37). Conclusions Implementation of the Keystone ICU project was associated with a significant decrease in hospital mortality in Michigan compared with the surrounding area. The project was not, however, sufficiently powered to show a significant difference in length of stay.

Standardised mortality rates are a poor measure of the quality of hospital care and should not be a trigger for public inquiries such as the investigation at the Mid Staffordshire hospital, say Richard Lilford and Peter Pronovost

Objectives To evaluate the extent to which intensive care units participating in the initial Keystone ICU project sustained reductions in rates of catheter related bloodstream infections. Design Collaborative cohort study to implement and evaluate interventions to improve patients’ safety. Setting Intensive care units predominantly in Michigan, USA. Intervention Conceptual model aimed at improving clinicians’ use of five evidence based recommendations to reduce rates of catheter related bloodstream infections rates, with measurement and feedback of infection rates. During the sustainability period, intensive care unit teams were instructed to integrate this intervention into staff orientation, collect monthly data from hospital infection control staff, and report infection rates to appropriate stakeholders. Main outcome measures Quarterly rate of catheter related bloodstream infections per 1000 catheter days during the sustainability period (19-36 months after implementation of the intervention). Results Ninety (87%) of the original 103 intensive care units participated, reporting 1532 intensive care unit months of data and 300 310 catheter days during the sustainability period. The mean and median rates of catheter related bloodstream infection decreased from 7.7 and 2.7 (interquartile range 0.6-4.8) at baseline to 1.3 and 0 (0-2.4) at 16-18 months and to 1.1 and 0 (0.0-1.2) at 34-36 months post-implementation. Multilevel regression analysis showed that incidence rate ratios decreased from 0.68 (95% confidence interval 0.53 to 0.88) at 0-3 months to 0.38 (0.26 to 0.56) at 16-18 months and 0.34 (0.24-0.48) at 34-36 months post-implementation. During the sustainability period, the mean bloodstream infection rate did not significantly change from the initial 18 month post-implementation period (−1%, 95% confidence interval −9% to 7%). Conclusions The reduced rates of catheter related bloodstream infection achieved in the initial 18 month post-implementation period were sustained for an additional 18 months as participating intensive care units integrated the intervention into practice. Broad use of this intervention with achievement of similar results could substantially reduce the morbidity and costs associated with catheter related bloodstream infections.

Changes that can improve patients’ health are often difficult to get into practice, even when backed by good evidence. Peter Pronovost, Sean Berenholtz, and Dale Needham describe a collaborative model that has been shown to work