Article Text

Download PDFPDF

How reliable is your hospital? A qualitative framework for analysing reliability levels
  1. David E Ikkersheim,
  2. Marc Berg
  1. Plexus, Straatweg 68, Breukelen, The Netherlands
  1. Correspondence to Dr David E Ikkersheim, Delft University of Technology, IPSE studies, Plexus, Straatweg 68, 3621 BR Breukelen, The Netherlands; ikkersheim{at}plexus.nl

Abstract

Objective Many approaches and methods have been developed to reduce errors in the healthcare delivery process and to increase patient safety. One of the approaches suited to improve patient safety is reliability theory. This paper adds a qualitative dimension to the application of reliability theory in hospitals. Based on a review of the literature, the authors identified a framework of qualitative elements that can be used to diagnose, understand and thereby improve upon the level of reliability in (department of) a hospital.

Results Based on the literature search, the authors identified four interconnected elements that are crucial for hospital reliability. These four elements are: process optimisation and standardisation; outcome measurement and monitoring; responsibilities and accountability of medical professionals; and organisational culture.

Discussion Substantial effort has been made in the last decade to improve patient safety. The actual improvement in safety has been fairly modest, which is understandable because most hospitals currently have fairly unreliable processes in place. Using the framework presented here, hospitals can gauge the reliability of their processes and practices. Recognisable characteristics provide insights into where improvement is needed and possible. In addition, this framework provides a way to view the relationship between different patient safety building blocks and a means to link them conceptually. An integrated approach is needed for hospitals to achieve a higher reliability level with particular attention to the interconnected elements that affect patient safety.

  • Patient safety
  • reliability
  • human error
  • quality of care
View Full Text

Statistics from Altmetric.com

Introduction

Ever since the report ‘To err is human’ was published in 1999, patient safety has been high on the agenda of healthcare professionals, managers and policy makers. Many approaches and methods have been developed to reduce errors in the healthcare delivery process, thereby decreasing adverse events and increasing patient safety.1

One of the approaches suited to improve patient safety is reliability theory. This theory has been successfully applied in high-reliability sectors such as the nuclear-power and air-traffic-control sectors. Like the healthcare sector, these are sectors where work is done at the interface between the social (human behavioural) and technical components of complex systems.2–4

A high-reliability organisation is an organisation that is extremely well focused on preventing failure, on expecting the unexpected and on ensuring that the errors that unavoidably will occur will not result in catastrophic events.5–7As we will see, high-reliability organisations stem from, among others, high-reliability processes.

The reliability of a process is operationalised by determining the number of defects that occur for every 10 opportunities. The unit of measurement for reliability is therefore 10−x (see table 1). The definition of the ‘process’ obviously influences the reliability level that will be found. For example, the process from intake to discharge (including medication, diagnostics, surgery, etc) of a patient may be taken as one process. But one could also consider the process of administering medication as a separate process. For comparing practices, it is crucial to use the same definitions between these practices.

Table 1

Process reliability levels5

In comparing sectors, reliability studies have invariably described healthcare delivery processes as ‘chaotic’ (ie, less than 10−1 reliability).5 8 9

This paper adds a qualitative, evolutionary dimension to the concept of organisational reliability in a healthcare organisation. Based on a review of the literature, we identified a framework of qualitative elements that can be used to indicate and improve the level of reliability in a hospital. Managers and professionals can determine the reliability level of their hospital or department and determine if their hospital should be on the left or right side of the boxes shown in figure 1.

Figure 1

Average risk in various industries and activities of fatal outcome per process initiation (adapted from Amalberti et al).8 The reliability levels in this study are higher than in table 1 because process failure here is expressed in terms of ‘fatal outcome’.

Methods

For an overview of methods, practices and concepts that could be used to improve the reliability of hospitals, we conducted a literature review (on 21 April 2010). The syntax entered into PubMed was: ‘patient safety AND reliability.’ We screened the abstracts of 899 papers for relevance, and selected 41 pertinent papers that were read in greater detail. The 34 papers that described a valid method, practice or concept that improved the reliability of hospital processes were included in our framework. In addition, we screened key papers on this topic and checked references of included papers. Only Dutch and English studies were included.

Analysis

First, we categorised the results based on how the presented method, practice or concept improved patient safety. We labelled each study with the most appropriate key words and then focused on the commonalities between the studies based on these key words. Subsequently, we found that the core categories revolved around four interconnected elements, which we subsequently developed further on the basis of the papers investigated. This step led to the development of the evolutionary model described below.

Results

We identified four interconnected elements that are crucial for hospital reliability. Figure 2 shows these four elements: process optimisation and standardisation; outcome measurement and monitoring; responsibilities and accountability of medical professionals; and organisational culture.

Figure 2

Four interconnected elements that influence the reliability level of a hospital.

These four elements are discussed below; specifically, we describe what hospital characteristics are associated with low- and high-reliability processes for each element. Our findings are summarised in table 2, where we distinguish the common characteristics for each level of reliability using the scale presented in table 1.

Table 2

Reliability framework based on a literature review and analysis

Process optimisation and standardisation

The potential gain in reliability by standardisation is visible in medical activities such as anaesthesia and blood transfusion, where the failure rates are orders of magnitudes lower than in other parts of medical work (see figure 1). Historically, in these areas of medicine, processes have become more standardised, and professionals fulfil their tasks as a streamlined part of standardised, sociotechnical routines. To achieve comparable failure rates, standardisation is essential for many other medical processes as well. Achieving such a goal implies articulating the goals of the process and carefully mapping the processes involved.5 Processes need to be stripped of unnecessary variation and complexity using protocols, checklists, decision aids and assisted with automated reminders generated by electronic medical records.5 10–15 If processes are always performed in concordance with each other, bundling these processes to ensure that all processes are properly executed further improves reliability.16 Reliable processes are designed in such a way that the default processes are the right processes; forms, organisational routines and/or material constraints structure the process in such a way as to ensure optimal safety.17 Once this is achieved, adding failure identification and failure mitigation can create additional redundancy. This results in ‘fail-safe’ processes, thereby further improving reliability.18 19

Outcome measurement and monitoring

A first step towards reliable healthcare is to register outcomes, especially adverse outcomes.17 20 21 Hospitals begin by registering adverse events (such as pressure ulcus rates) and reporting incidents at a local (department) level and informing professionals about their performance.22 At baseline, there is no focus on performance benchmarking; at best there might be some attention on (mediocre) outcomes elsewhere. More reliable hospitals register these outcomes centrally over the full cycle of care across departments and include them in the planning and control cycle. The next step is benchmarking with other hospitals, a process facilitated by information technology; this facilitates rapid-cycle improvement of processes and aids in root-cause analyses of accidents.5 21 23 An even higher level of reliability can be reached with permanent benchmarking with best-in-class and the simultaneous use of statistical process control.5

Responsibilities and accountability of medical professionals

A classic hospital is organised in departments in which individual physicians care for patients. That is to say, physicians are individually responsible for the quality of care delivered, and their accountability to management or to other professionals in the hospital is limited.2 To improve the quality and reliability of processes, it is crucial that physicians, nurses and other professionals start working together in teams clustered around medical conditions.24 By working and training together in a dedicated team, communication failures are drastically reduced, and group performance levels—in routine and non-routine situations—improve significantly.11 19 25 26 The clinical leadership of these teams is accountable to colleagues and central management for the results the team achieves.2 27 Reaching higher levels of reliability, such teams become increasingly competent (‘self-steering’) at handling unexpected situations and dealing with the inevitable errors that will occur. This means that centralised patient safety routines can be delegated increasingly to the ‘front-line,’ thereby increasing the responsiveness of the organisations.11 28 Highly reliable teams possess skills such as rule-based decision-making and task allocation.19 29 Where possible, these teams use rules and/or checklists to make decisions regarding diagnosis and treatment of patients. Tasks regarding care for patients are allocated in such a way that they are conducted by people who are most suited to doing them, rather than on hierarchy. This results in an environment where mechanisms are in place to migrate decisions to those most suitable for the task regardless of their rank.6 These teams have situational awareness: they have a constant awareness of what is happening around them, and a constant focus on events that may happen and about which they cannot be fully informed.25 29

Organisational culture

Culture is a crucial element in achieving reliable processes in healthcare. Healthcare processes that are not reliable are often accompanied by a reactive working culture in which there is a denial of vulnerability (the ‘iron man mentality’) along with ‘blaming and shaming’ with no explicit focus on patient safety.8 30 Concepts for educating young professionals such as ‘see one, do one, teach one’ are prevalent in unreliable organisations, and new technologies are introduced without proper training of personnel. In reliable organisations, leaders instill a zero-tolerance culture in which patient safety is everyone's responsibility; there is a sense of the bigger picture of patient safety and related issues.21 31–35 In this environment, leaders view safety problems as problems with their system rather than with their employees.32 Proper training of personnel and application of a system-wide approach to risk and safety management by the leadership results in prioritisation of permanent improvement of quality and patient safety. All professionals within such an organisation recognise that their individual behaviour can improve patient and that a safe environment is not an exogenous condition which is only created by the leadership of the hospital.36 37 In a reliable healthcare culture, the mentality of ‘intent, vigilance, and hard work’ is abandoned. People realise that only a systematic approach can improve patient safety and that trying to work longer hours will not resolve fundamental issues which create chaotic processes. In this environment, critique of personnel is appraised and used to continuously create better processes and procedures.5

In high-reliability organisations, we find resilience and a focus on what went right and why, instead of a focus only on what went wrong. This means that there is an awareness why and how the hospital is successful in managing patient safety and what the factors are which create the safe environment. By knowing that, the factors which create the safe environment are ‘exported’ as best practices to other processes and departments across the hospital, thereby improving the reliability of the hospital. At the same time, in ultrasafe environments, there is a permanent awareness of things that might go wrong, a so-called ‘preoccupation with failure’ and ‘collective mindfulness.’ This means that people on the work floor know that inevitably things will go wrong, even in processes that are well organised. This collective awareness of front-line healthcare teams is then used to permanently monitor processes on possible failure. By doing so, the errors that occur are recognised in an early stage and are mitigated in such a way that fatal outcomes are avoided.6 38 39

Synthesis of the literature: linking concepts into a reliability framework

The elements described above, which collectively determine the reliability level of a hospital, are summarised in table 2. The presented framework distinguishes four phases of reliability per element. When hospitals make their processes more reliable, they gradually move from phase 0 until they finally reach phase 3.

Although the four presented elements are interconnected, hospitals do not per se have to be in the same phase for the four different elements. Different departments within the same hospital may be in different phases. In addition, it is also possible that a particular department is in phase 1, for example, the element ‘process optimisation and standardisation’, and is in phase 2 for the element ‘outcome measurement and monitoring.’

Discussion

Using the framework presented here, hospitals can gauge the reliability of their processes and practices. Recognisable characteristics provide insights into where improvement is needed and possible. In addition, this framework provides a way to view the relationship between different patient safety issues and a way to link them conceptually.28 Recognising the state of reliability within their hospitals, professionals and managers can use the presented framework as a starting-point to implement these concepts, methods and practices that suit their specific situation best.

This framework differs from other quality and safety frameworks such as the Baldrige and ISO frameworks as it specifically focuses on aspects of healthcare processes at the front line, rather than on competences of organisations regarding quality-management systems, strategic planning, etc.40 41 It also distinguishes itself from previous relevant frameworks, as it not only links the level of organisational reliability with process and cultural characteristics of an organisation, but also distinguishes four interconnected elements, with per element a detailed description of tangible process characteristics.2 5 6 In addition, it describes the evolutionary aspects of these elements by defining characteristics per phase. However, the feasibility, efficiency and effectiveness of this new framework require further evaluation and validation through empirical research.

It seems that hospitals around the world are still mostly in reliability phase 0 or 1. Progress in patient safety has been modest despite many patient safety- and quality-related initiatives.42 Improving reliability requires improvement in all four interconnected elements, and it seems difficult for hospitals to make progress in all these interconnected elements at the same time. Some areas of medicine such as anaesthesia and blood transfusion have already shown that it is possible to achieve a significantly higher reliability level by standardising and optimising processes. As high reliability on an organisational level primarily stems from highly reliable processes, only creating a safe culture will not substantially improve reliability. In line with previous studies, we find that culture seems to be the result from an organisation's shared values and its structure, practices and processes it has in place. Because shared values and beliefs are so interconnected with the ‘being’ of the organisation, it is almost impossible to change that independently, as culture seems to be an ‘emergent property.’43 (We thank one of the anonymous reviewers for suggesting the term ‘emergent property’ and thereby strengthening our conclusions).

Therefore, the focus of hospitals should be on standardising processes, measuring and monitoring outcomes as well as focusing on medical professionals' responsibilities and accountability. By doing so, the needed cultural change will ‘emerge,’ which could be reinforced by giving special attention to the cultural change. After making healthcare processes reliable, hospitals can start with implementing more sophisticated high-reliability organisation concepts such as resilience and situational awareness.5

In conclusion, substantial effort has been made in the last decade to improve patient safety. The actual improvement in safety has been fairly modest, which is understandable because most hospitals currently still have phase 0 or phase 1 processes in place. An integrated approach is needed for hospitals to achieve a higher reliability level. Given the current state of patient safety in most hospitals, particular attention to the characteristics within the interconnected elements that affect patient safety at the process level is needed to become more reliable than they are now.

References

View Abstract

Footnotes

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.