Background

In health accreditation a standard is “a desired and achievable level of performance against which actual performance is measured” [1]. Standards enable “health service organisations, large and small, to embed practical and effective quality improvement and patient safety initiatives into their daily operations” [2]. External organisational and clinical accreditation standards are considered necessary to promote high quality, reliable and safe products and services [2, 3]. There are over 70 national healthcare accreditation agencies worldwide that develop or apply standards, or both, specifically for health services and organisations [4].

The International Society for Quality in Health Care (ISQua) seeks to guide and standardise the development of these agencies and the standards they implement [5]. ISQua advocates that accreditation standards themselves need to meet exacting standards, and has standards for how to develop, write and apply them. ISQua conducts the International Accreditation Program (IAP) for the certification or accreditation of standards against their standards [5]. The International Standards Organisation (ISO), a network of the national standards institutes of 162 countries, is the largest developer and publisher of international standards [6]. Standards from ISO are also applied in international health jurisdictions.

In short, healthcare standards, and standards for standards, are ubiquitous. They are advocated to be an important means of improving clinical practice and organisational performance. ISQua, and many national bodies, espouse, and have documented methodologies to promote open, transparent, inclusive development processes where standards are developed by members [6–11]. They assert that their methodologies are effective and efficient at producing standards appropriate for the health industry. However, the evidence to support these claims requires scrutiny. What is the basis to ground the standard development methodologies in use? What research demonstrates how standards should be crafted and structured to ensure they are understandable, unambiguous, achievable and reliable in making assessments? What studies have identified the necessary steps to enable standards to be incorporated into everyday practice? Is there evidence to show whether standards improve practice? The purpose of this study was to examine these questions by identifying and analysing the research literature focusing on the development methods and application of healthcare accreditation standards.

The analysis is a systematic narrative synthesis of the literature [12]. The intention is to generate new insights and bring transparency to the topic under investigation [13, 14]. This type of review is appropriate for this topic for four reasons. First, the review aims to examine a complex initiative applied in diverse contexts [15]. That is, accreditation programs are complex organisational interventions, trying to shape both organisational and clinical conduct, within a multifaceted context in turn shaped by, for example, the healthcare and policy environment. Second, accreditation programs, involving healthcare standards, have been researched in different ways by divergent groups. The analysis method adopted here is intended specifically for interventions researched in a myriad of ways [12]. Third, the approach enables consideration of apparently disparate data generated by research into accreditation standards, as a complex organisational intervention [15]. Fourth, the questions being investigated are preliminary questions that need to be asked of this intervention and the approach is designed exactly for this [14, 15]. The review differs from previous reviews [16, 17] in being specifically focused only on healthcare accreditation standards and not the broader “standards” field. This review is the first to undertake a systematic and detailed narrative synthesis of accreditation standards.

Methods

Selection criteria and search strategy

The selection criteria were: peer-reviewed, publicly available English language empirical research papers on the topic of healthcare accreditation standards. Discussion and commentary, and non-English language papers were excluded. Despite these focused criteria, we recognise that they may capture heterogeneous literature including, possibly, an overlap with work covering other forms of regulation. To counter this potential problem we used a staged search strategy to identify and remove any papers not focused on the study topic. This approach is valid for two reasons. First, there are overlaps between how regulatory strategies are at times discussed in the literature [18–20]. The reviewing of abstracts or the full papers provided a mechanism by which to screen out literature not on the study topic. Second, previous reviews and a preliminary investigation signalled that empirical research literature available on standards was limited.

A multi-method strategy based on similar review designs was employed [16, 21, 22]. There were three stages (see Figure 1). The search was first conducted in March 2010 and updated in August 2011. Citations and abstracts that met the search criteria were downloaded into Endnote X.0.2, a reference management program. Abstracts and, where uncertainty arose, complete papers, were reviewed against the selection criteria for inclusion in the review.

Figure 1
figure 1

Literature search, review and selection flow chart.

Full size image

The first stage had three steps. First, we selected databases in the health sector. Literature was drawn from five electronic bibliographic databases: Medline, Psych INFO, EMBASE and Social Work databases from 1980, and CINAHL (nursing and allied health literature) from 1982. Second, we identified abstracts focusing on the topic of ‘accreditation’. Third, we selected abstracts using the terms ‘standard’, ‘guideline’, ‘policy’ and ‘legislation’; where appropriate, terms were truncated with the symbol ‘$’ and searched using the ‘Exp’ function to capture widest publication of papers (for example, guideline$ or polic$). The initial search yielded 9386 abstracts (including duplicates). We reviewed the selection to exclude those not written in English and also to remove duplicates.

In the second stage we refined the collected abstracts. Two researchers independently reviewed the abstracts, selecting papers using two criteria. We selected for empirical research studies, using derivations of phrases such as ‘research’, ‘study’, ‘empirical’ or ‘report’, and ‘method’. Using this strategy the selection was reduced to 2111 articles. This group was further analysed to identify those papers that covered ‘impacts’ of accreditation standards. At this point we removed papers covering clinical or biomedical issues and also discussion pieces, commentaries or editorials. To supplement the formal search process, two less structured search methods were implemented. We undertook a 'snowballing' search, which is a variation on snowballing sampling [23]. That is, we examined the assembled manuscripts reference lists for additional relevant papers potentially missed in the formal search. In parallel, an investigation of websites of agencies associated with the study topic, that is, reports or papers investigating the evidence base for accreditation or quality standards in the health sector, was conducted. We searched: the ISQua research site; the websites of 31 healthcare accreditation agencies worldwide; ISO website; and standards organisations’ websites of a number of countries (Additional file 1: Appendix 1). The application of the stage 2 refinement processes to the collected abstracts yielded 140 articles.

In the third stage, to determine the final selection of papers meeting the study criteria, two experienced researchers independently reviewed the identified 140 papers and discussed their relevance. The focus was the selection of papers that addressed development methods and application of healthcare accreditation standards. This stage derived 13 articles.

Analysis

The selected papers were analysed by three independent researchers in two ways. First, the characteristics of the studies were noted. For each paper a summary of authors, country, sector, aim, methods, major findings and conclusions, and study quality was compiled. The level of evidence was assessed using Australian National Health and Medical Research Council guidelines [24] and study quality by an assessment tool developed from publically available checklists [21, 25]. Together they enabled examination of study quality, incorporating intervention or aetiology (that is, impact), level of evidence, design and appraisal of quality (Table 1). Second, a narrative analysis of the literature was conducted in line with the study aims.

Table 1 Quality rating assessment criteria*
Full size table

Results

The 13 papers were synthesised (Table 2). The results are presented under three headings: standards development; implementation issues; and the impact of standards. The papers were examined according to date of publication, country, sector, methodology and focus.

Table 2 Assessment of empirical healthcare standards research
Full size table

Study details, characteristics and quality

The dates of the studies ranged from 1995 to 2009 inclusive. The majority of studies were published in two years, 2003 [20, 26–29] and 2009 [17, 18, 30, 31], with five and four studies, respectively. One study was published in each of the following years: 1995 [32], 2004 [33], 2007 [34] and 2008 [35]. Studies were conducted in six countries. The United States of America (USA) was the setting for the majority of studies (n = 8) [17, 18, 26, 29, 31–34]. The remaining five countries all had one study: United Kingdom [35]; Philippines [30]; Australia [20]; South Africa [27]; and Taiwan [28]. The studies were all conducted in the acute sector (n = 13). The majority of studies had a multidisciplinary focus (n = 9) [17, 18, 20, 26–28, 32–34] and the practices of nurses [30, 35] and managers [29, 31] were the individual focus of two studies each. Research projects used mixed methods [20, 32–35], employed quantitative methodologies to examine archival databases [17, 18, 26, 28, 31] or undertook a questionnaire survey [27, 29, 30]. Within the mixed methods studies the qualitative tools were questionnaires, surveys, interviews, reviews and evaluations. The quantitative methods covered examination of databases, prospective and retrospective studies and stratified randomised studies. The study content was categorised according to the focus of the papers, that is, program, clinical or workplace issues. Program issues was the topic that most studies examined via four different program sub-topics: reviews of programs (n = 5) [18, 20, 28, 29, 31]; policy compliance (n = 4) [17, 32–34]; program impacts (n = 3) [26, 27, 30]; and organisational environment (n = 1) [35]. Just five studies had content relating to clinical care [17, 18, 20, 26, 34] and one on staff workplace issues [35].

A summary of the intervention or impact (aetiology) assessment, level of evidence classification and quality ratings for the selected literature is represented in Table 3. Using the NHMRC guidelines, three investigations [27, 32, 35] were classified as interventions and ten studies [20, 26, 28–33, 36] under the aetiology criteria. In the intervention group, Aiken et al. (2008), was assessed as meeting the fourth level of evidence and all the quality criteria. While Salmon et al. (2003) and Stradling et al. (2007) were rated at the second and fourth level of evidence rating, respectively, each were missing some study details and so were rated at the second level for quality ratings. The studies within the aetiology group were divided between the two top quality levels. Six [26, 29, 30, 32, 33, 36] were rated as meeting all criteria, and four [28, 29, 31, 37], while missing some but not significant information to compromise them, were rated on the second tier of quality.

Table 3 Summary of the intervention or aetiology assessment, level of evidence classification and quality ratings
Full size table

Standards development

No study directly examined standards development or other issues associated with their progression. That is, no empirical study was identified which examined: what is best practice for developing standards; standard development processes; the wording or structure of standards; or what types of standards would have the greatest likelihood of improving practice.

Implementation issues

Only one study examined implementation issues with healthcare accreditation standards [33]. Five factors were noted as assisting implementation: external pressure from legislation and accreditation; the use of technology and self-evaluation as tools to leverage change; organisational culture characteristics; research; and peer education. Conversely, three factors were reported to hinder implementation: lack of external incentives or pressure; organisational policies and culture; and cost and resource constraints [33].

Impact of standards

Twelve of the 13 papers addressed the impact of standards [26–32, 35–37]. The impact of the standards on the organisation, clinical quality and staff could be identified.

Impacts of standards on the organisation

The single randomised controlled trial identified demonstrated that compliance with accreditation standards increased in the intervention group, from 38 to 76%, compared to in the control group, from 37 to 38% [27]. Furthermore, standards or guidelines about the organisation of clinical practice led to improved efficiency and quality practices. Specifically, standards within an accreditation program resulted in decreased length of hospital stay [26], improved management of disclosure of preventable harm [29], and utilisation of patient safety practices [36].

Impacts of standards on clinical quality

Accreditation program standards encompassing trauma care [26], prenatal care [30], post partum care [37], stroke care [32], breastfeeding [28], pain management [29], and the institution wide organisation of care [27, 30, 33] were reported to improve the provision of care. Additionally, there were links to improvements in various aspects of clinical quality. For example, standards contributed to: reductions in in-hospital mortality and length of stay [26], and rates of infections and decubitus ulcers [36]; and improvements in breastfeeding rates [28] and the proportion of patients receiving relevant tests, medications and admission for stroke [32]. Conversely, and at times simultaneously, standards introduced to improve care appeared not to do so. For example, exposure to standards for prenatal and delivery care [30], document control [31], and the organisation of care [27] did not show any measurable effects. Nor did rates of certain adverse events, such as failure to rescue or postoperative respiratory failure, alter with the implementation of accreditation standards [36].

Impact on staff

Standards were shown to produce an improved staff quality of life, working conditions and appraisals of the quality of care. This outcome was noted from the use of ‘Magnet’ principles, which sought to improve the attraction of the workplace in recruiting and retaining staff [35]. Additionally, the introduction of standards, through an accreditation program, resulted in the improved perceptions of teamwork and participation in decision making [27], and compliance with tobacco control [32].

Discussion

This study employed systematic search procedures to academic databases and accreditation agency websites to uncover empirical research that grounds the development methods and application of healthcare accreditation standards. The review has built on previous work in the healthcare accreditation field [16, 17], commencing where previous reviews finished. We started with a proposition that standards are ubiquitous within healthcare and are generally considered to be an important means by which to improve clinical practice and organisational performance. However, the evidence about whether accreditation standards change behaviour of health care organisations, clinical quality and staff is at best equivocal, and is determined by the circumstances.

Only three intervention studies were identified in the review. Two interventions resulted in improvements attributed to the implementation of accreditation standards [32, 35]. The improvements were the organisational working environment and staff perceptions [35], and care processes and appropriateness of care [32]. The remaining study, conducted in a developing country [27], involved health services seeking improvement from a very low base and hence the applicability of the results is limited to that context. The non-intervention studies have shown that, whilst there is adherence to standards in some cases, in a range of instances there is little evidence as to their effects. In short, the effectiveness of the development, writing, implementation and impacts of healthcare standards are significant issues that lack convincing evidence.

It is not clear, for example, what might be evidence-based practice in the development of standards. However, the literature synthesis suggests that reoccurring strategies include mobilising external leverage, organising teams or creating receptive cultures within health care organisations to optimise the opportunity to create standards. Yet an overarching finding was that applying standards has mixed results. There is limited published peer-reviewed evidence regarding the correspondence between the application of standards and improvements in organisational performance, clinical quality or staff behaviours.

There is the opportunity for the standards development field to learn from the experience of people developing technical standards, practice guidelines and evidence-based clinical policies. Consideration can be given to the applicability of translation of development processes and implementation strategies from other areas in healthcare [38–40]. The Joint Commission in the USA, for example, through the establishment of the National Patient Safety Goals initiative has used development and implementation processes from which lessons can be learnt [41].

Agencies setting standards, including accreditation bodies or programs that develop or apply them, or both, also have significant experience and expertise in conducting these activities. Some have been doing so for decades. More recently, ISQua is utilising and sharing this experience through two strategies: the ISQua IAP and the accreditation workshop conducted at ISQua’s annual international quality conference. The ISQua IAP has been implemented to “build credibility and comparability for national organisations by harmonising standards and procedures on common international principles” [42]:349. Established in 1999, the IAP utilises the expertise of senior people within accreditation agencies to review, offer ideas for improvements, and accredit programs in other countries. ISQua reports that the IAP has accredited 19 organisations and 35 sets of standards (from 21 organisations), and eight surveyor training programs [36]. Each year the accreditation workshop at the ISQua international quality conference draws together practitioners and researchers from around the world to consider current developments and challenges associated with healthcare accreditation programs [43]. Discussions have centred upon all aspects of accreditation programs, for example: implementation of accreditation programs [44, 45]; maintaining standards of accreditation programs [46]; survey methodologies [47, 48]; linking standards to clinical indicators [49]; processes used to develop standards [50]; and the public disclose of accreditation results [51, 52].

Conclusion

The challenge is to translate practical experiences and discussions into rigorous empirical evidence. We lack knowledge of how to strengthen the development of standards and the application of them based on sound critically peer-reviewed evidence. The process to develop standards essentially needs to be transformed from learnt experience to a verifiable, evidence-based methodology. Evidence-based mechanisms by which standards are developed, promulgated, reinforced, audited and evaluated are needed. Linking the writing of standards, including the wording, structure, design, focus and content, to improved outcomes requires further rigorous investigation. Factors that promote or inhibit implementation of standards, and the impacts that result, need detailed examination and analysis. This review has revealed some significant gaps in our knowledge in these areas, and, in doing so, extended previous reviews in the healthcare accreditation field.

As to the limitations of our study, while we have endeavoured to be systematic, we may have overlooked some important literature. A further limitation is that papers or reports needed to be publicly available and in English to be included in the results.