Health Policy 123 (2019) 661–665

Contents lists available at ScienceDirect

Health Policy
journal homepage: www.elsevier.com/locate/healthpol

Health service accreditation stimulating change in clinical care and

human resource management processes: A study of 311 Australian
hospitals
David Greenfield a,∗ , S.A. Lawrence b , A Kellner b , K. Townsend b , A. Wilkinson b
a
Australian Institute of Health Service Management, University of Tasmania, NSW, Australia
b
Centre for Work, Organisation and Wellbeing, Griffith University, QLD, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Objective: This study aimed to establish whether longitudinal participation in an accreditation program
Received 11 December 2018 is translated into improvement in continuity of quality patient care and human resource management
Received in revised form 7 April 2019 (HRM) processes outcomes.
Accepted 24 April 2019
Materials and methods: This was a secondary data analysis of accreditation panel data from acute hospitals
participating in the Australian Council on Healthcare Standards’ Evaluation and Quality Improvement
Keywords:
Program (EQuIP). EQuIP criteria data from 311 hospitals were collected by external surveyors across
Accreditation
2003–2006 (Time 1) and 2007–2010 (Time 2). Mandatory accreditation criteria ratings at Time 1 were
Quality improvement
Measurement of quality
used to determine hospital performance group membership (1 = below moderate, 2 = moderate, 3 = above
Patient outcomes moderate). Analysis was undertaken of ratings across continuity of quality patient care and HRM process
Human resources criteria, at Time 1 and 2.
Process change Results: Continuity of quality patient care and HRM processes improved across time in the three per-
formance groups. Lower performing hospitals improved at a greater rate than moderate and higher
performing hospitals. The groupings and performance order did not change over time.
Conclusions: An accreditation program is an external driver that facilitates continual and systemic quality
improvement changes to sub-systems with an organisation.
© 2019 Elsevier B.V. All rights reserved.

1. Introduction grams, in developing and developed settings, are an important

external driver to improve the safety, quality and effectiveness of
External healthcare regulation bodies, such as Accreditation organisational and clinical processes [8–11].
Canada, The Australian Council of Healthcare Standards (ACHS) Performance in an accreditation program is reflected in higher
and Haute Autorité de Santé, or government Ministries, for exam- levels of clinical performance outcomes [11–14], the effectiveness
ple, the Lebanese Ministry of Health, implement accreditation of organisational systems, including human resource management
programs which promote organisational compliance and improve- (HRM) systems [12,15] or organisational care processes [3,11].
ment against standards. Programs typically do this through a Determining if participation in an accreditation program across
longitudinal and incremental approach to modifying management several cycles results in continual process improvements, com-
and care processes and subsystems [1,2]. The perceived high pliance to external standards or both, continues to be points of
cost and time investment institutions commit to accreditation concern for stakeholders, including policy makers and govern-
programs, in developing and developed countries, means some ments [3,10,16,17]. Participation in an accreditation program, in
stakeholders continually question the contribution they make to a developing country, the United Arab Emirates, was shown to
improving care outcomes and organisational functioning [3–7]. stimulate and maintain organisational improvements in quality
Nevertheless, there is sufficient evidence that accreditation pro- measures within an accreditation cycle [3]. Similarly, in Australia,
a developed country, it was demonstrated that the effective func-
tioning of an organisation’s HRM system can be enhanced through
participation in an accreditation cycle; the caveat being that the
∗ Corresponding author at: University of Tasmania, Locked Bag 5052, Alexandria, motivation for excellence is internal so that the accreditation pro-
New South Wales, 2015, Australia. gram reinforces, rather than drives, performance improvements
E-mail address: david.greenfield@utas.edu.au (D. Greenfield).

https://doi.org/10.1016/j.healthpol.2019.04.006
0168-8510/© 2019 Elsevier B.V. All rights reserved.
662 D. Greenfield et al. / Health Policy 123 (2019) 661–665

[15]. Whether improvements are maintained across cycles, or are Participating organisations are required to undertake actions to
realised in other settings, is not known. self-assess and improve their performance against the accredita-
The associations between accreditation and organisational- tion standards. This includes focusing on both organisational and
level quality of patient care or process outcomes have normally clinical systems and processes [11,14]. An organisation assesses
been examined at single points in time. We have limited empirical itself against the EQuiP Standards and produces a self-assessment
evidence of the longitudinal impact of participation in an accredi- report for the AHCS. This report is reviewed by an external peer-
tation program on organisational performance outcomes. One case survey team which also conducts an on-site visit to verify the
study that has taken a longitudinal focus demonstrated that one improvement claims, documentation and care practices. The sur-
hospital, in a developing country, did sustain improvements across vey team provides a written report back to ACHS, which can
the accreditation program cycle [3]. Further investigations are nec- include recommendations for further improvement and the grant-
essary to reveal if this finding holds for other organisations, in both ing of accreditation status. Within this four year period, usually
developing and developed countries. This is an issue with signifi- at mid-point or thereabouts, a further onsite survey assessment is
cant policy, financial and, quality and safety implications. undertaken. A surveyor team visits to corroborate the continued
A key question concerning many stakeholders is: does the qual- achievement of the safety and quality standards, implementation
ity improvement component of an accreditation program, across of any recommendations and confirm ongoing accreditation status
several cycles, translate into identifiable organisational outcomes? [21–23].
That is, across accreditation cycles, does organisational perfor- External peer-surveyors, employed by ACHS, visited and rated
mance improve, stall or decrease? Furthermore, for organisations hospitals against the EQuIP standards criteria across two accred-
with different levels of accreditation performance, do they demon- itation periods. As per the ACHS accreditation programs normal
strate the same or different levels of improvement longitudinally? practice, different survey teams visited each hospital in each
The unique contribution of this study was to investigate these period. Where possible at least one member of the survey team
issues through focusing upon continuity of quality patient care was retained for the subsequent visit. Previous research into
and HRM processes. This study aimed to establish whether lon- the accrediting agency’s surveyor program has demonstrated the
gitudinal participation in an accreditation program is translated strategies and processes that promote reliability [24]. This includes
into improvement in continuity of quality patient care and HRM the requirement that surveyors attend annual training address-
processes outcomes. ing knowledge developments and current interpretation directions
associated with the program standards and assessing practices.
Survey teams were also recognised as an important mediating
2. Method influence on individuals, promoting consistency in interpretation
and reliability in an accreditation program [24]. Nevertheless, in an
2.1. Program setting, sample and research design attempt to account for variations in individual surveyor and survey
team assessments the study population of hospitals were classi-
The study focused on the ACHS accreditation program, the Eval- fied into one of three mandatory accreditation performance groups,
uation and Quality Improvement Program (EQuIP), implemented in which became the focus of analysis.
Australia [18]. At the time of the study, accreditation of acute and The first assessment period was between 2003-6, and used
sub-acute care organisations was a policy endorsed by State Gov- EQuIP3 (Time 1) and the second was between 2007-10, and used
ernments and the ACHS program accredited facilities from all States EQuIP4 (Time 2). The EQuIP standards are rated on a five-point
and Territories across Australia. ACHS was the major accreditation scale, from ‘1’ indicating low achievement, ‘3’ corresponding to
agency for acute and sub-acute care organisations, with over 1300 moderate achievement to ‘5’ representing outstanding achieve-
facilities accredited, representing over 90% market share [19]. ACHS ment. Inclusion criteria for study participants were that they were
has been facilitating an accreditation program for 40 years, com- accredited through the ACHS accreditation program, from the pub-
mencing in 1974; organisations now have participated in several lic or private acute hospital sector, for both time periods. The study
accreditation cycles, including under EQuIP [20]. used participating organisations’ EQuIP mandatory standards cri-
ACHS uses the approach of responsive regulation [21]. That teria outcomes as secondary panel data for analysis.
is, engaging with industry representatives to develop, implement
and revise EQuIP; the development process is incremental with 2.2. Measures and analysis
revisions approximately every three years to incorporate develop-
ments in the safety and quality knowledge base. The adjustments Following the processes established by Townsend et al. the
from editions three to four were to increase the focus on consumer study measures and analysis processes were implemented [1,2].
participation and the need for evidence of clinical and organisa- First, three hospital specific details were used as control variables
tional outcomes [19]. EQuIP, based on the principle of continuous in the analysis: ownership (1 = public, 2 = private); geographical
quality improvement, operates on a four-year cycle and is divided regions (state/territory regions were binary coded 0 = hospital not
into three broad assessment categories: clinical, support and cor- in this geographic region, 1 = hospital in this geographical region);
porate criteria [19]. Across these categories there are 13 standards and hospital size (1 = 1–49 beds, 2 = 50–99 beds, 3 = 100–199 beds,
with 45 criteria in total - 14 mandatory and 31 non-mandatory 4 = 200–499 beds, 5 = more than 500 beds).
items. The breakdown is as follows: clinical category, six standards Second, mandatory accreditation performance groups were
and 21 criteria - seven mandatory, 14 non-mandatory; support derived. Data were analysed to classify participating organisa-
category, five standards and 14 criteria - three mandatory, 11 non- tions into one of three mandatory accreditation performance groups:
mandatory; and, corporate category, two standards and 10 criteria below moderate (1), moderate (2) and above moderate (3). The
- four mandatory, six non-mandatory. An organisation’s quality groupings were achieved by calculating an average mandatory
and safety achievements, and efforts to implement improvement accreditation performance score, for each hospital at Time 1; this is
strategies, are rated against a five-point scale (Little Achievement, the first assessment for each hospital and different for each [12,13].
Some Achievement, Moderate Achievement, Extensive Achieve- Finally, two further measures were derived from the accredi-
ment and Outstanding Achievement). Ratings of at least Moderate tation data: an HRM processes score and a continuity of quality of
Achievement against the mandatory criteria are necessary to obtain patient care score, which were composites of five and six items,
accreditation status. respectively. Organisational means for each score were calculated
 D. Greenfield et al. / Health Policy 123 (2019) 661–665 663

Table 1
Repeated measures analysis of covariance results for dependent variables.

Variable HRM processes Continuity of quality patient care

Type III sum of F Partial eta Type III sum of F Partial eta df
squares squared squares squared

Within-subjects effects
Time .000 .007 .000 .016 .383 .001 1, 302
Time X mandatory accreditation performance groups 2.062 18.767*** .111 1.741 20.755*** .121 2, 302
Between-subjects effects
Mandatory accreditation performance groups 2.773 42.927*** .221 2.615 47.873*** .241 2, 302

Note: Analyses are controlling for hospital sector, geographical region and size.***p < .001.

by transforming the EQuIP standards ratings from their five-point

word scale into a numerical scale, that is, for example, from low
achievement to ‘1’, moderate achievement to ‘3’, and, outstand-
ing achievement to ‘5’. A unit-weighted mean composite score
for each was then developed [2]. The HRM processes score com-
bined five accreditation criteria items: HRM planning; recruitment
and selection; training and development; performance manage-
ment; and, support services. The continuity of quality of patient care
score, reflecting the hospital’s provision of a seamless process of
quality patient care, was constructed using six items: prioritised
access to care according to clinical need; assessment identifies cur-
rent and ongoing need; consumer/patient are fully informed and
provide consent; best possible care is planned and delivered in part-
nership with consumer/patient; care is evaluated together with
consumer/patient; and, ongoing care needs are addressed during
discharge/transfer).
Additionally, as construct and reliability analysis for the two
measures had not previously been established, confirmatory factor Fig. 1. Simple slope effects for Continuity of quality patient care.
analyses, as detailed by Townsend et al. and based on estab-
lished processes was undertaken [1]. ANCOVA repeated-measures
were conducted. Time was entered as the within-group variable,
After controlling for variance in the dependent variable
mandatory accreditation performance group membership as the
accounted for by sector, geographical region and size, the test of
between-group variable, and control variables (hospital size, sec-
within-subjects effects revealed a significant interaction of manda-
tor, geographical region) as covariates. Fisher’ “protected t” (LSD)
tory accreditation performance groups and time (T1 to T2) on the
procedure was used to guard against Type 1 errors [25]. The study
Continuity of quality patient care (F (2, 302) = 20.755, p < .001; par-
focus was on the population of organisations undergoing accredita-
tial eta squared = .12) (Table 1). Analysis revealed significant simple
tion, not specific institutions. Improvements at the level individual
effects for all three mandatory accreditation performance groups:
of organisations, due to unrecognised internal changes or surveyor
group 1 - below moderate (F (1, 302) = 110.483, p < .001; partial eta
team inconsistency, could not be accounted for.
squared = .268); group 2 - moderate (F (1, 302) = 30.814, p < .001;
partial eta squared = .093); and group 3 - above moderate (F (1,
3. Results 302) = 7.805, p < .01; partial eta squared = .025) (Fig. 1).
After including controls, the test of within-subjects effects
There was a population of 311 hospitals who participated in the revealed a significant interaction of mandatory accreditation perfor-
accreditation process across both cycles, Time 1 and Time 2. The mance groups and time (T1 to T2) on HRM processes (F (2, 32) = 18.77,
population included organisations from all Australian States and p < .001; partial eta squared = .11). Analysis revealed significant
Territories and was nearly evenly split between the public (53%) simple effects for all three mandatory accreditation performance
and private (47%) sectors. The distribution of hospitals across the groups: group 1 - below moderate (F (1, 302) = 101.073, p < .001;
size categories was as follows: 23% (1), 25% (2), 25% (3), 20% (4) partial eta squared = .251); group 2 - moderate (F (1, 302) = 21.206,
and 7% (5). The distribution of values across these demographic p < .001; partial eta squared = .066); and group 3 – above moderate
variables is representative of those of the ACHS hospital population (F (1, 302) = 9.250, p < .01; partial eta squared = .030) (Fig. 2).
at that time (N = 483). The tests of within-subjects effects account for the individual
The results for HRM processes score at Time 1 and Time 2 differences within each organisations’ level of performance. Hence,
were: T1 - Mean = 2.92, SD = .29, Cronbach’s alpha = .77; and T2 - the results, for both continuity of quality patient care and HRM pro-
Mean = 3.10, SD = .26, Cronbach’s alpha = .70. Similarly, the continu- cesses scores, indicate significant changes, or improvements, for
ity of quality patient care score at Time 1 and Time 2 were: T1 - these measures in the groupings over time, that is, from T1 to T2.
Mean = 3.00, SD = .25, Cronbach’s alpha = .76; and T2 - Mean = 3.17, In sum, the continuity of quality patient care and HRM pro-
SD = .25, Cronbach’s alpha = .73. cesses scores significantly improved over time for all three of the
These results indicate acceptable measures of internal consis- mandatory accreditation performance groups. Additionally, manda-
tency for both the HRM processes and continuity of patient care tory accreditation performance group membership influenced the
scores, that is, the findings demonstrate that the two composite relative strength of improvement over time: improvement was
scores are reliable measures. For reasons of parsimony, only the greatest for those organisations in the below moderate group, less
results for statistically significant interactions of interest are illus- so for the moderate group and slightly further reduced for the above
trated in Table 1. All other interaction effects are non-significant. moderate group.
664 D. Greenfield et al. / Health Policy 123 (2019) 661–665

demonstrated that participation in an accreditation program:

maintained and stimulated improvements in quality measures
across cycles [3]; enhanced HRM performance within a cycle [15];
contributed to improvements in collective quality improvement
action [28]; and, stimulated quality-related policy and strategic
planning [29]. As a result two important impacts are identified.
First, individual organisations and the health system they are
embedded within accrue system efficiency improvements through
participation in an accreditation program. Second, patient care
quality systems, processes and measures improve from accred-
itation assessments. Therefore, a key policy question is: should
participation in an accreditation program now be a mandatory
requirement by government?
This longitudinal study is an important contribution to
strengthening the accreditation evidence base for programs, and
methodologically this marks it out as different from much of the
previous work. Future research investigating other accreditation
programs or clinical performance measures will be useful to deter-
mine if similar results are realised. Additionally, this work can
Fig. 2. Simple slope effects for HRM processes.
be extended through studies examining an accreditation program
with data from three, or more, cycles thereby allowing for the use
of random coefficient modelling to further interrogate the longitu-
4. Discussion dinal relationships and impacts.
The data used in the study came from all organisations within
Organisational improvement in continuity of quality patient care the AHCS program, and as this cohort represents the overwhelming
and HRM processes outcomes, as represented by improved scores majority of hospitals in Australia, there was no selection bias. Sim-
against accreditation standards criteria, could be clearly identified ilarly, as all organisations were subject to the same national policy
in all three groups, across two accreditation survey cycles. Organ- context and developments across the study period, any impacts are
isational performance, on these criteria at least, did not stall or assumed to be consistent for the cohort. However, we do acknowl-
decrease, but clearly shifted in a positive improvement direction. edge the possible impact of a differential effect based on earlier
This is important evidence to address known stakeholder concerns performance. That is, the higher performing organisations faced
[10,16,17]. a ceiling to their possible improvement based on their previous
Organisations commencing with different levels of accreditation scores. While this effect is present, what the findings show is that all
performance showed different rates of improvement against the organisations continued to improve while remaining consistently
standards criteria. Organisations initially demonstrating the low- placed relative to each other’s performance. The top performing
est performance against the standards, relative to their peers, were hospitals remained the in the top sub-cohort and the others sim-
shown to improve at greater rates. The improvement rates of the ilarly placed in relation to them; organisations did not change
middle group also exceeded that of the top group. Additionally, groupings across the sub-cohorts. Hence the effect, while present,
the performance gap between the three groups narrowed signifi- does not negate the importance of the study findings.
cantly across the two cycles, albeit differently for the two outcomes
examined. Both performance interval differences at Time 2 were 5. Conclusion
significantly reduced compared to Time 1. However, just as sig-
nificantly, the performance order of the three groups did not alter Longitudinal participation in an accreditation program trans-
across the two cycles. That is, the higher performing organisations lated into evidence of ongoing compliance with and performance
at Time 1 remained the top group at Time 2, and, similarly, the improvement against external standards. This is one further piece
middle and bottom groups at Time 1 occupied the same positions of evidence that accreditation programs, in developing and devel-
at Time 2. Furthermore, across different standards criteria the rates oped settings, are an important external driver to improve the
of improvement for the different groups of organisations mirrored safety, quality and effectiveness of organisational and clinical pro-
each other. That is, for example, the rate change for the lowest group cesses. Not surprisingly, as healthcare organisations are recognised
for both items investigated is nearly identical. This suggests that an as complex adaptive systems, research in this field is revealing
organisation’s improvement from one cycle to the next might be accreditation programs have a multifarious and interwoven impact
the same for across all standard criteria. across the systems and processes of organisations.
Taking a ‘big picture’ view of the study and considering the three
groups collectively, enables us to see quality and safety of the acute
care sector, as measured by performance against the accredita- Acknowledgements
tion standards, improved over the two cycles. Participation in the
accreditation program, across several cycles, contributed to com- This research was supported under the Australian Research
pliance to external standards and continual process improvements. Council Linkage Projects scheme (project numbers LP120100325
This is one piece of evidence that validates claims to this effect, and LP100200586).
made by accreditation bodies.
The findings are consistent with research that has highlighted References
the potential for accreditation to facilitate continual and sys-
tematic quality improvement changes to sub-systems within [1] Townsend K, Wilkinson A. Managing under pressure: HRM in hospitals. Human
Resource Management Journal 2010;20:332–8.
hospitals [26,27]. The knowledge base recording the positive [2] Townsend K, Lawrence SA, Wilkinson A. The role of hospitals’ HRM in shaping
impact of accreditation programs continues to grow. Studies across clinical performance: a holistic approach. International Journal of Research in
developed and developing country settings, and over time, have Human Resource Management 2013:1–24.
 D. Greenfield et al. / Health Policy 123 (2019) 661–665 665

[3] Devkaran S, O’Farrell PN. The impact of hospital accreditation on quality agement: lessons from an Australian study. International Journal for Quality in
measures: an interrupted time series analysis. BMC Health Services Research Health Care 2014;26:372–7.
2015;15:137. [16] Greenfield D, Braithwaite J, Pawsey M. Health care accreditation surveyor styles
[4] Mitchell JI, Izad Shenas SA, Kuziemsky C. Governance standards: a roadmap typology. International Journal for Quality in Health Care 2008;21:435–43.
for increasing safety at care transitions. Healthcare Management Forum [17] Touati N, Pomey M-P. Accreditation at a crossroads: are we on the right track?
2015;28:28–33. Health Policy 2009;90:156–65.
[5] Mumford V, Forde K, Greenfield D, Hinchcliff R, Braithwaite J. Health services [18] Australian Council on Healthcare Standards. National report on health services
accreditation: what is the evidence that the benefits justify the costs? Interna- accreditation performance 2003-2006. Sydney: Australian Council of Health-
tional Journal for Quality in Health Care 2013;25:606–20. care Standards; 2007.
[6] Saleh SS, Bou Sleiman J, Dagher D, Sbeit H, Natafgi N. Accreditation of hospi- [19] ACHS. 2008-2009 annual report. Ultimo, NSW: The Australian Council on
tals in Lebanon: is it a worthy investment? International Journal for Quality in Healthcare Standards; 2009.
Health Care 2013;25:284–90. [20] ACHS. Background on accreditation. ACHS; 2018 https://www.achs.org.au/
[7] Shaw C. Accreditation is not a stand-alone solution. Eastern Mediterranean about-us.
Health Journal 2015;21:226–31. [21] Greenfield D, Pawsey M, Braithwaite J. Accreditation: a global regulatory
[8] Greenfield D, Pawsey M, Braithwaite J. What motivates professionals to engage mechanism to promote quality and safety. In: Sollecito W, Johnson J, editors.
in the accreditation of healthcare organizations? International Journal for Qual- Continuous quality improvement in health care. New York: Jones and Barlett
ity in Health Care 2011;23:8–14. Learning; 2013. p. 513–31.
[9] Australian Commission on Safety and Quality in Health Care. Annual report [22] Greenfield D, Hinchcliff R, Westbrook M, Jones D, Low L, Johnston B, et al.
2012-13. Sydney: Australian Commission on Safety and Quality in Healthcare; An empirical test of accreditation patient journey surveys: randomised trial.
2013. International Journal for Quality in Health Care 2012;24:495–500.
[10] Hinchcliff R, Greenfield D, Moldovan M, Westbrook JI, Pawsey M, Mumford V, [23] Greenfield D, Moldovan M, Westbrook M, Jones D, Low L, Johnston B, et al. An
et al. Narrative synthesis of health service accreditation literature. BMJ Quality empirical test of short notice surveys in two accreditation programs. Interna-
& Safety 2012;21:979–91. tional Journal for Quality in Health Care 2012;24:65–71.
[11] Alkhenizan A, Shaw C. Impact of accreditation on the quality of healthcare [24] Greenfield D, Pawsey M, Naylor J, Braithwaite J. Are accreditation surveys reli-
services: a systematic review of the literature. Annals of Saudi Medicine able? International Journal for Quality in Health Care 2009;22:105–16.
2011;31:407–16. [25] Cohen J, Cohen P. Applied multiple regression/correlation for the behavioral
[12] Braithwaite J, Greenfield D, Westbrook J, Pawsey M, Westbrook M, Gibberd R, sciences. Hillsdale: Erlbaum; 1983.
et al. Health service accreditation as a predictor of clinical and organisational [26] Greenfield D, Braithwaite J. Health sector accreditation research: a systematic
performance: a blinded, random, stratified study. Quality & Safety in Health review. International Journal for Quality in Health Care 2008;20:172–83.
Care 2010;19:14–21. [27] Pomey M-P, Contandriopoulos A-P, François P, Bertrand D. Accreditation: a
[13] Miller MR, Pronovost P, Donithan M, Zeger S, Zhan C, Morlock L, et al. Rela- tool for organizational change in hospitals? International Journal for Quality in
tionship between performance measurement and accreditation: implications Health Care 2004;17:113–24.
for quality of care and patient safety. American Journal of Medical Quality [28] Desveaux L, Mitchell JI, Shaw J, Ivers NM. Understanding the impact of accred-
2005;20:239–52. itation on quality in healthcare: a grounded theory approach. International
[14] Falstie-Jensen AM, Nørgaard M, Hollnagel E, Larsson H, Johnsen SP. Is com- Journal for Quality in Health Care 2017;29:941–7.
pliance with hospital accreditation associated with length of stay and acute [29] Berssaneti FT, Saut AM, Moreno MC. Evaluating the impact of accreditation on
readmission? A Danish nationwide population-based study. International Jour- Brazilian healthcare organizations: a quantitative study. International Journal
nal for Quality in Health Care 2015;27:450–7. for Quality in Health Care 2017;29:713–21.
[15] Greenfield D, Kellner A, Townsend K, Wilkinson A, Lawrence SA. Health service
accreditation reinforces a mindset of high-performance human resource man-