Incentives for Health Care Providers: An Overview of Reviews

Transcription

1 Incentives for Health Care Providers: An Overview of Reviews Workforce Research & Evaluation Health Professions Strategy & Practice Alberta Health Services Research Team: Renee Torgerson, Jana Lait, Gail D Armitage, Jordana Linder, Shelanne Hepp, Karen Jackson, Esther Suter August 8 th, 2012

2 Table of Contents SOURCES OF SUPPORT... 1 DECLARATION OF INTEREST... 1 ACKNOWLEDGEMENTS... 1 EXECUTIVE SUMMARY... 2 INTRODUCTION... 4 RESULTS Description of the Health Workforce and Workplace Setting Incentives and HHR Outcomes Effectiveness of Financial Incentives Effectiveness of Non-Financial Incentives Incentives for Rural and Remote Areas Incentives for New Graduates DISCUSSION Effectiveness of Financial Incentives Effectiveness of Non-Financial Incentives Effectiveness of Incentives for Certain Provider Groups, Health Care Contexts and Demographics IMPLICATIONS FOR POLICY AND RESEARCH REFERENCES Tables Table 1: Methodological quality of included reviews by review... 9 Table 2: Characteristics of Included Table 3: Outcomes & Incentives Table 4: Outcomes & Incentives for Rural and Remote Areas Table 5: Outcomes & Incentives for New Graduates Figures Figure 1: Process of reviewing abstracts and reviews... 9 Figure 2: Methodological quality of included reviews by AMSTAR criteria... 10

3 Appendices Appendix A: Search parameters Appendix B: Abstract screening instructions Appendix C: AMSTAR tool Appendix D: References for Included Reviews Appendix E: References for Excluded Reviews... 46

4 SOURCES OF SUPPORT Funding provided by the Institute of Health Economics in a request for work directed to the Institute for Public Health at the University of Calgary. The following authors are members of the Institute for Public Health at the University of Calgary: Dr Esther Suter and Dr Renee Torgerson This report is based on research funded by the Institute of Health Economics under the auspices of Alberta Health. The views expressed herein do not necessarily represent the official policy of the Institute of Health Economics or Alberta Health. DECLARATION OF INTEREST The authors report no conflict of interest. ACKNOWLEDGEMENTS The research team would like to acknowledge the following individuals for their contributions to the overview: Ms. Betty-Lynn Morrice, Associate Chief Health Professions Officer and Vice President, Health Professions Strategy and Practice, Alberta Health Services Ms. Brenda Rebman, Vice President, Human Resources, Alberta Health Services, Alberta Health Services Dr. Herb Emery, Svare Professor in Health Economics, University of Calgary Ms. Elizabeth Aitken, Librarian, Health Information Network, Rockyview General Hospital Knowledge Centre, Alberta Health Services Workforce Research and Evaluation, Alberta Health Services 1

5 EXECUTIVE SUMMARY Background A perceived shortage and a misdistribution of health care providers has prompted much consideration about how to best get the right number of health care providers with the right skills and qualifications to the right places and how to retain them in the long-term. Incentive strategies are aimed at improving overall health human resource (HHR) outcomes, however, there is a lack of information on what incentives are most effective to recruit and retain health care providers. Objectives The objective of this overview is to identify and assess evidence-based financial and nonfinancial incentives for improving HHR outcomes for a range of health care providers within different workplace contexts and settings. This overview assesses and summarizes reviews that report on incentives being utilized in health care systems similar to Canada. It is intended as a resource for Canadian employers for designing an evidence-based incentive strategy. Methods This overview constitutes a review of systematic reviews published in 2000 or later that have summarized effectiveness of incentives for health care provider recruitment and retention. This methodology is an efficient approach for reviewing, contrasting, comparing, and synthesizing a range of variables. The literature search produced 416 reviews that were screened for inclusion in our overview. Physician remuneration models were out of scope for this overview and were excluded. Thirty-three reviews met the inclusion criteria and were evaluated for methodological quality by at least two research members using the AMSTAR tool. Of those, 13 reviews met the quality criteria and were included in the overview. Information from these 13 reviews was extracted including a description of the review, the incentives reported on and the impact on HHR outcomes. The information on the relationship between incentives and HHR outcomes was assessed and synthesized. Results and Conclusions The overview reports on the following HHR outcomes: recruitment, retention, job satisfaction, absenteeism, turnover and intent to leave/continue to practice. Incentives were separated into financial and non-financial incentives, with non-financial incentives categorized as follows: Positive work environments (workload, autonomy, clinical and social support, and work-life balance); Workforce Research and Evaluation, Alberta Health Services 2

6 Supports for career and professional development (promotional opportunities, clinical supervision, and educational programs); Work design (staffing models, employment status, and collaboration) Incentives for the following two contextual and demographic groups were reported on by four of the reviews: Incentives for rural and remote areas Incentives for new graduates None of the incentives emerged as a magic bullet for improving HHR outcomes. The effectiveness of incentives varied with most reviews producing mixed evidence. Given the available evidence, it is likely that a combination of financial and non-financial incentives is needed to successfully recruit and retain health care providers. The following general conclusions about the effectiveness of incentives can be made: 1. There is inconclusive evidence for the effectiveness of financial compensation as a sole mechanism for improving HHR outcomes. There is some evidence that both direct and indirect financial compensation are important elements within an incentives package for recruiting health care providers, especially physicians, to rural/remote communities. 2. There is supporting evidence that improving the workplace environment, especially the promotion of work life balance, is key for improving HHR outcomes. However, there is inconclusive evidence for the effectiveness of incentives related to workload (e.g. job demands and task complexity) and the promotion of autonomy. These incentives require further review. 3. Providing supports for professional development and career opportunities including clinical supervision and education programs are effective for improving job satisfaction and recruitment of health care providers and may be particularly effective for new graduates. 4. Few reviews outlined any targeted incentives specific to professional groups, health care and employment contexts or demographics. Workforce Research and Evaluation, Alberta Health Services 3

7 INTRODUCTION Discussions on health workforce issues have intensified in recent years due to the predicted shortage of health care providers across Canada and noted areas of misdistribution (e.g. rural and remote areas). Health care policy makers and employers have the challenge of attracting health care providers, deploying them where they are most needed, and retaining them in the long-term (Stillwell, 2011). The health workforce crisis is attributed to a convergence of factors including an aging population, a rise in the complexity of diseases requiring more intense resources, and an aging health workforce (Dieleman & Harnmeijer, 2006; Mathauer & Imhoff, 2006; Spence Laschinger et al., 2009; World Health Organization, 2006). In Canada, discussions on workforce planning have centred on workforce management models aimed at balancing workforce supply and demand, and deploying available human resources and skill mix to optimize workforce productivity and efficiency (Birch et al., 2007; Purdy, et al., 2010; Hall, Doran & Pink, 2008). Part of addressing workforce challenges involves examining issues around health human resources (HHR). HHR has been described as the human capital needed to design health care systems and to implement health service delivery models that are cost effective (Health Canada, 2011). Often focusing on the provision of an adequate supply and appropriate mix of health care providers ideally based on population health needs (Advisory Committee on Health Delivery and Health Human Resources [ACHDHR], 2007), the ultimate goal of planners is to optimize HHR outcomes including retention, recruitment, job satisfaction, motivation, intent to leave, absenteeism, engagement and productivity. Improving HHR outcomes requires a wide range of incentives, which include, all the rewards and punishments that providers face as a consequence of the organizations in which they work, the institutions under which they operate and the specific interventions they provide (World Health Organization, 2000 p.61). Incentives constitute some form of payment or benefit that achieves a specific change in behaviour or influences a preference for one choice over other alternatives such as influencing a person s decision about not only entering the health workforce, but also where to work and practice (Flogren et al., 2011; Buchan, Thompson & O May, 2000). There are two main types of incentives, that is, financial and non-financial incentives. Financial incentives encompass all direct (e.g. salaries and wages) or indirect compensations (e.g. scholarships and bonuses) (Kingma, 2003; International Council of Nurses, 2008). While financial incentives are typically the first line in employment negotiations, on their own they may not be sufficient for improving HHR outcomes (Buchan, Thompson & O May, 2000). They often need to be complemented by a range of non-financial incentives to achieve desired HHR outcomes, including incentives that relate to positive work environments (Adams & Hicks, 2000; Shamian & El-Jardali, 2007), supports for career and professional development (Adams & Hicks, 2000) and work design (e.g. staffing models and skill mix). Workforce Research and Evaluation, Alberta Health Services 4

8 Designing effective incentives for an engaged, satisfied and productive workforce is highly complex and incentives range considerably in magnitude and type. The effectiveness of incentives is influenced by the existing labour market, employment context, and organization of professional groups (such as union coverage) amongst other factors. Moreover, what is effective for one professional group in a particular setting might not have the same draw for other professions and other contexts. Employers may also face challenges related to the availability of both human and financial resources to develop incentives to meet their labour force needs. For example, drawing people to less desirable areas may require higher compensation packages which may in turn mean trade-offs in other areas of expenditure. There is likely no one size fits all approach for recruiting and retaining health care providers (Fooks et al., 2002; McIntosh, Torgerson & Wortsman, 2007; ACHDHR, 2007). Rather, incentives need to be aligned with local and organizational needs, the availability and composition of the workforce, available resources, and desired outcomes. The objective of this research is to summarize the evidence on effectiveness of incentives for improving HHR outcomes in health care. The focus is on the effects of incentives on HHR outcomes in general, however, other incentives pertinent to different provider groups, demographics and employment contexts are also considered. For example, attracting people to relocate to and stay in rural and remote regions presents considerable challenges across Canada (Humphreys et al., 2012; Dolea, Stormont & Braichet, 2010; Grobler et al., 2009) and therefore, incentives such as bursaries and higher salaries are often used (Bärnighausen & Bloom, 2009; Van Diepen, McRae & Paterson, 2007, Frehywot et al., 2010). Likewise, designing targeted incentives for different health care provider groups and age cohorts may be useful, recognizing their unique contractual obligations and labour market realities. For instance, providers working in the private sector (e.g. private practice psychologists) may have different expectations and require a different set of incentives than providers working in the public sector. There are also specific challenges related to improving HHR outcomes for providers from different age cohorts and levels of experience (e.g. new graduates versus providers nearing retirement; Laschinger et al., 2009; McGillis Hall et al., 2011; Salt, Cummings & Profetto-McGrath, 2008; Wortsman & Janowitz, 2006). METHODS An overview of systematic reviews of the effects of financial and non-financial incentives on HHR outcomes was conducted. Overviews of systematic reviews are an efficient way of summarizing at a high level the evidence in an area of interest (Thomson et al., 2010). Workforce Research and Evaluation, Alberta Health Services 5

9 Research Questions The following research questions guided the overview: 1. What financial and non-financial incentives have been considered for improving HHR outcomes? 2. Which of these financial and non-financial incentives are the most effective for improving HHR outcomes? 3. Are there any incentives that are particularly effective for certain provider groups and health care contexts (for example, urban vs. rural/remote, unionized vs. non-unionized health care environment, publicly vs. privately funded), and demographics (e.g. age cohorts)? Criteria for Review Inclusion Any type of review article (Cochrane reviews, systematic reviews, narrative reviews) was considered for inclusion in this overview as long as the review met the following criteria: Identification of regulated or unregulated health care providers English or French language Reviews relating to HHR systems in New Zealand, Australia, United Kingdom, United States, Sweden, Denmark, France, Germany, Finland, Norway or Canada Reviews published between Inclusion of one or more financial or non-financial incentive (see search strategy Appendix A; remuneration models for physicians such as fee-for-service, capitation, and mixed payment models were excluded) Inclusion of one or more HHR outcomes (see search strategy Appendix A; patient outcomes, quality assurance, service delivery or quality of care were excluded) Peer-reviewed articles (dissertations, thesis, grey literature, editorials were excluded due to time limitations) Reviews including single studies with any type of study design Meet minimum quality standard (scores 4 points or higher on AMSTAR rating) Search Strategy for Identification of The research team developed the search terms and search parameters in consultation with an Alberta Health Services human resources vice president responsible for recruitment and retention and a research librarian proficient with health care databases who executed the search strategy (See Appendix A for the full search strategy). The following electronic health databases were searched: MEDLINE (Ovid) Embase (comprised of Excerpta Medica Database, Drugs and Pharmacology, Psychiatry) CINAHL Workforce Research and Evaluation, Alberta Health Services 6

10 Cochrane Database of Reviews PsycINFO Evidence Based Medicine Reviews (comprised of ACP Journal Club, Cochrane Central Register of Controlled Trials, Cochrane Database of Reviews, Cochrane Methodology Register, Database of Abstracts of Reviews of Effects, Health Technology Assessment, NHS Economic Evaluation Database) The following electronic business databases were searched: ABI Inform Trade & Industry Business Source Complete A hand search of the Health Systems Evidence repository of the McMaster Health Forum was also conducted. Assessment Process The assessment process for inclusion was multi-staged. Stage 1 Abstract retrieval and review Abstracts were downloaded into a reference database and duplicates removed. Each abstract was independently screened by three readers for eligibility. Differences in agreement amongst readers regarding inclusion were resolved through consensus. See Appendix B for the abstract screening instructions and eligibility criteria. Stage 2 Review retrieval and eligibility screening Reviews which met the inclusion criteria based on the abstract screening were retrieved. These reviews were read by two researchers to determine eligibility using the same inclusion criteria as used during the abstract screening (Stage 1). Reading of the full reviews led to the exclusion of a number of reviews as they did not meet inclusion criteria. The remaining reviews advanced to Stage 3. There are two articles that report on the same review (Humphreys et al 2009; Buykx et al 2010). Differences in agreement between researchers were resolved through consensus. Stage 3 Methodological quality screening Reviews which met the inclusion criteria during Stage 2 were read independently by two researchers to determine methodological quality using the AMSTAR tool (Assessment of Multiple Systematic Reviews) and to confirm eligibility based on inclusion criteria used during Stages 1 and 2. AMSTAR is a validated tool for assessing the methodological quality of systematic reviews and was developed by clinicians and epidemiologists (Shea et al., 2007; Shea et al., 2009). The AMSTAR is comprised of 11 criteria. Each statement was scored yes, no, can t answer or not applicable. Differences in agreement between the two researchers regarding the AMSTAR scores were resolved by discussion to reach consensus. A score of 1 was assigned to each yes response; a zero score was assigned to all others. Workforce Research and Evaluation, Alberta Health Services 7

11 Reviews scored 0-4 indicate low quality and were excluded from further consideration. Reviews receiving AMSTAR scores of five and higher were moderate or high quality and were included in this overview. See Appendix C for the AMSTAR tool. Data extraction Two reviewers sequentially extracted data from the included reviews using an extraction sheet designed for this overview. The extraction sheet captured details of each review s aim, details of search, the number and designs of studies included in the review, the countries represented, the providers, settings (e.g. acute care, primary care), and main conclusions. The second extractor had access to the first extractor s data and validated and complemented the information as needed. Summaries of Incentives Data extractions were the basis of summaries of the incentives. The summaries focused on a general description of the health workforce and workplace setting discussed in the reviews, followed by a summary of effectiveness of financial and different non-financial incentives. Special sections were dedicated to incentives for rural and remote areas and incentives for new graduates. RESULTS A total of 416 abstracts were screened; three hundred and fifty-two abstracts were judged to not meet the inclusion criteria and were excluded from further consideration for this overview, leaving 64 reviews to advance to Stage 2. During stage 2, 31 reviews were excluded and 33 advanced to stage 3, the AMSTAR rating process. During this stage, another 20 reviews were excluded based on poor methodological quality rating (n=6) or ineligibility because they did not meet inclusion criteria after scrutiny (n=14), leaving 13 reviews for extraction with two reviews reporting on the same report (Humphreys et al 2009; Buykx et al 2010). Thus there are 13 reviews from 14 reports included in the study. Figure 1 displays the number of abstracts or reviews at each stage. A list of all included and excluded reviews is in Appendix D and E respectively. Workforce Research and Evaluation, Alberta Health Services 8

12 Figure 1: Process of reviewing abstracts and reviews Note: Two included reviews report on the same study. Thus there are 13 reviews from 14 reports included in the overview. Table 1 shows the AMSTAR scores assigned to each of the 13 studies. The quality of the included studies ranged from achieving 6 of the 11 quality criteria to 10 of the 11 criteria. Reviews rated high on a priori study design, literature searches, characteristics of studies provided, and the ways the studies were assessed and used to formulate conclusions (Figure 2). Table 1: Methodological quality of included reviews AMSTAR criteria (see Appendix C for AMSTAR tool and statements) Author Total YES Butler 2011 Y Y Y Y Y Y Y Y Y N N 9 Buykx 2010/Humphreys 2009 Y Y Y Y N Y Y N N N Y 7 Chenoweth 2010 Y CA Y Y N N Y Y N N Y 6 Davey 2009 Y N Y Y N Y Y Y N Y N 7 Workforce Research and Evaluation, Alberta Health Services 9

13 Dolea 2010 Y CA Y Y N CA CA Y N Y Y 6 Hodgkinson 2011 Y Y Y Y Y Y Y Y Y N Y 10 Hunter 2002 Y N N Y Y Y N N N Y N 5 Lee 2008 Y N Y Y N Y Y Y N Y N 7 Patterson 2010 Y N Y Y N Y CA CA Y Y N 6 Salt 2008 Y N N Y N Y Y Y N Y N 6 van Ham 2006 Y N Y N N Y Y Y N N N 5 van Wyk 2010 Y Y Y N Y Y Y Y Y N Y 9 Zangaro 2007 Y Y Y Y N Y Y Y Y Y N 9 Figure 2: Methodological quality of included reviews by AMSTAR criteria The following section describes the findings from the reviews included in our overview. It should be noted that differences between short- medium- and long-term outcomes were not discussed in the literature. As well, while some reviews included individual studies that reported quantitative results, others did not; thus, it was not possible to do a synthesis of quantitative results. Table 2 displays the data extraction table that was used to describe the characteristics of the reviews and analyze the information. Workforce Research and Evaluation, Alberta Health Services 10

14 Table 2: Characteristics of Included Reviews First author, year Butler, 2011 Buykx, 2010 (& Humphreys, 2009) Chenoweth 2010 Aim of review Explore how hospital nursing staff models affect patient and staff outcomes Effectiveness of retention strategies for health workers in rural/remote areas in Australia Identify evidencebased factors supporting recruitment and retention of nurses in aged and dementia care Details of search Inception Databases not reported databases databases Description of studies Total : 15 Include: RCT (n=8) Controlled clinical trials (n=2) Controlled before/after (n=5) Total :14 Include: evaluations (n = 6),review articles (n = 8) Total : 25 Include: not reported Countries represented Corresponding Author: Ireland include: Australia, Netherlands, USA, Canada, England Corresponding Author: Australia include: USA, Australia, Canada, Japan, New Zealand, Africa (may include other international but does not explicitly state which countries) Corresponding Author: Australia Include: Australia, UK, USA, Sweden, Canada, Finland (may Description of providers in studies Hospital nursing staff (RNs or equivalent, licensed practical or equivalent, unlicensed assistive personnel or equivalent) Doctors, nurses, multidisciplinary, hospital administrators, allied health in rural/remote areas Dementia and aged care nurses (registered, enrolled, licensed, practical, director of nursing, advanced practice) Type of settings in studies Hospitals (acute, non acute, all sizes, teaching and non-teaching, public and private Not reported Aged care settings (acute, residential, community) Main Results Interventions relating to hospital nurse staffing models may improve some staff-related outcomes. In particular, primary nursing and self scheduling may reduce staff turnovers Financial incentives are common retention strategies; however, there is inconclusive evidence that they are the most important factor in the decision to remain in rural/remote areas. The authors found evidence that nonfinancial incentives such as adequate relief, autonomy and child care/family supports influence retention. There is supporting evidence for strategies that bundle retention incentives, rather than relying on one strategy alone. Developing a family-friendly, learning environment that values its staff is important for recruiting and retaining nursing staff. Some successful nurse recruitment and retention strategies include: careful selection of student nurse clinical placements and ongoing supervision and education; training for new and existing nurses; increased autonomy; pay parity across health settings; and family friendly policies. It is more effective to combine these factors rather than relying on a single incentive. Workforce Research and Evaluation, Alberta Health Services 11

15 First author, year Davey, 2009 Dolea, 2010 Hodgkinson, 2011 Aim of review Identify and examine predictors of short-term absences of staff nurses working in hospital settings Effectiveness of intervention strategies to attract and retain health workers in remote and rural areas Identify controlled trials that evaluate nursing models in residential aged care designed to Details of search databases September databases, websites Various databases Description of studies Total : 14 Include: Prospective studies (n=7), Others not reported (n=7) Total : 27 Include: Longitudinal cohort (n=5), Retrospective (n=3), Pre and post (n=5), Control (n=1), Crosssectional observation (n=12) Total : 2 Include: Interrupted time series (n=1), Countries represented include other international but does not explicitly state which countries) Corresponding Author: Canada Include: USA, Canada, Netherlands, Israel, others not reported Corresponding Author: Switzerland Include: Australia, Canada, Japan, New Zealand, USA, Africa, Latin America, south-eastern Asia Corresponding Author: Australia Include: Canada, Netherlands Description of providers in studies Nurses (registered, licensed practical, ICU, general, new) Health providers in rural/remote areas, developed and developing countries, all types of health providers Nurses (registered and enrolled), personal care attendants Type of settings in studies Acute care hospitals Rural/remote settings (not specified) Aged care settings (residential, sub-acute, and extended) Main Results Nurses prior attendance records were the best predictor of absenteeism. Work attitudes (job satisfaction, organizational commitment, job involvement), pay and retention factors (e.g., promotional opportunity) provided inconclusive and mixed results in relation to nurse absenteeism. Rurally oriented medical education programmes, professional support and financial incentives (e.g., allowance, scholarships/loan repayment programs) positively influence the decision of graduates to practise in rural areas. There is some evidence to suggest that a primary-care model is preferred by staff over team nursing; however, there were no significant improvements in staff outcomes (i.e., job satisfaction, absenteeism, turnover). Workforce Research and Evaluation, Alberta Health Services 12

16 First author, year Hunter, 2002 Lee, 2008 Patterson, 2010 Aim of review reduce negative and improve positive outcomes for residents Assess the evidence on the influence of continuing professional development on enhancing the recruitment and retention of occupational therapists A systematic review of studies that examine determinants of front line nurse managers job satisfaction To assess the evidence on the relationship between human resource management (HRM) and individual and organizational outcomes Details of search Search years not reported 3 databases, 1 bulletin databases, 4 journals, 7 websites Up to June databases Description of studies Controlled before-andafter (n=1) Total : 13 Include: Not reported Total : 14 Include: Quantitative (n=12), Mixed methods (n=1), Qualitative (n=1) Total : 10 Include: Experimental (n=2), Prospective (n=8) Countries represented Corresponding Author: Not Reported Include: USA, UK Corresponding Author: Canada Include: USA, Canada, Hong Kong, UK Corresponding Author: United Kingdom Include: UK, USA, Denmark, Sweden, Netherlands Description of providers in studies Occupational therapists, clinical specialists Frontline nurse managers (nurses in leadership roles responsible for managing a nursing unit or team, and having direct supervision of staff nurses in a healthcare organization) Type of settings in studies Hospitals, Not reported Healthcare organizations Main Results Recruitment and retention of occupational therapists is influenced by a combination of personal and professional factors (e.g., autonomy, opportunities to develop skills, salary). There is insufficient evidence to conclude that continuing professional development alone increases recruitment and retention. Strategies that address autonomy, workload, increasing organizational support for managers and empowering managers to participate in decision-making may improve nurse manager s job satisfaction. Health professionals Not reported Human Resource Management methods can be used to support organisational change. Work design practices that enhance employee autonomy and control positively influenced job satisfaction, absence and health. Involving employees in the design/implementation of changes that affect their work is an effective strategy for enhancing HHR outcomes including job satisfaction. Incorporating training to support implementing change was emphasized. Workforce Research and Evaluation, Alberta Health Services 13

17 First author, year Salt, 2008 Van Ham, 2006 van Wyk, 2010 Aim of review Systematic review on the effectiveness of organization retention strategies for new graduate nurses Determine what factors influence job satisfaction of general practitioners Determine the effects of preventative staff-support interventions for health workers Details of search ( ) 11 databases, 8 research web sites, reference lists of key selected articles databases, reference lists Varied databases Description of studies Total : 16 Include: Onetime experimental case study design (n=9),onegroup pretestposttest design (n=3), Nonrandomiz ed control group pretestposttest design (n= 3), Static group comparison design(n=1) Total : 24 Include: Not Reported Total studies: 10 Include: RCTs Countries represented Corresponding Author: Canada Include: USA Corresponding Author: Not Reported Include: Not Reported Corresponding Author: South Africa Include: USA, Canada, Taiwan, Japan, others not Description of providers in studies New graduate nurses (NGNs), registered nurse residents, registered nurse interns, registered nurse externship, preceptors, managers, patient care directors, experienced nurses Type of settings in studies Hospitals Main Results Four types of retention strategies were implemented to increase retention rates for new graduate nurses (NGN). Implementing a retention strategy is effective for increasing retention rates of NGNs. The most successful retention rates were linked to a preceptor program model with an NGN focus and a program length of 3 to 6 months, while strategies implemented for less than 3 months were least successful. General practitioners Not Reported Factors that both increase and decrease job satisfaction were reported. The main factors that increased job satisfaction include diversity of work, relations and contact with colleagues, and being involved in teaching medical students. The main factors that decreased job satisfaction related to employment conditions and include low income, working hours and workload, and lack of recognition. Nurses of various ranks (from nurse managers to nursing aides), mixture of health professionals, healthcare team as a unit Tertiary, secondary, various settings including tertiary, secondary, residential, community, referral There is strong evidence to support the effectiveness of an intensive, long-term stress management training intervention on reducing job stress and risk of burnout among a wide range of health workers in various settings. Stress management training interventions should include periodic refresher sessions up to 18 months post-intervention to maintain beneficial effects of the training beyond the intervention. However, insufficient evidence was found to conclude that stress management training interventions positively affected job satisfaction and absenteeism over the Workforce Research and Evaluation, Alberta Health Services 14

18 First author, year Zangaro, 2007 Aim of review To examine the strength of the relationship between autonomy, job stress, and nursephysician collaboration on jobsatisfaction among registered nurses Details of search databases Description of studies Total : 31 Include: Not reported Countries represented reported. Corresponding Author: USA Include: England, USA, Canada, Australia, Taiwan, Sweden, Israel, Hong Kong, Scotland, United Kingdom Description of providers in studies Staff nurses Type of settings in studies Hospitals, specialty facility, multiple sites Main Results short and medium term. Job satisfaction is improved by facilitating good collaboration between nurses and physicians, and by fostering autonomy in the workplace. Workforce Research and Evaluation, Alberta Health Services 15

19 1. Description of the Health Workforce and Workplace Setting Of the 5 reviews that reported on the workplace setting, 3 focused on health care providers in hospitals and acute care settings (Butler et al., 2011; Davey et al., 2009; Salt, Cummings & Profetto-McGrath, 2008) with the remaining 2 reviews focusing on health care providers in continuing care (Chenoweth et al., 2010; Hodgkinson et al., 2009). No reviews focusing on home care, public health or primary and community care settings were identified in our search. Three of the reviews included a wide range of health care providers (Buykx et al., 2010; Dolea, Stormont & Braichet, 2010; Patterson et al., 2010). Some like Dolea and colleagues (2010) did not define the types of health care providers included in the review, making it more difficult to make any specific comments about the type of incentives specific to different provider groups. Seven reviews focused on nurses from different nursing groups (e.g. registered nurses, licensed practical nurses) and at different stages of their careers (Butler et al., 2011; Chenoweth et al., 2010; Davey et al., 2009; Hodkinson et al., 2011; Lee & Cummings, 2008; Salt, Cummings & Profetto-McGrath, 2008; Zangaro & Soeken, 2007). Finally, one review focused on general practitioners (Van Ham et al., 2006), and one review focused on occupational therapists (Hunter & Nicol, 2002). 2. Incentives and HHR Outcomes For the purpose of this review, we used the following categories to report on HHR outcomes as they were deemed most relevant by our HHR research team members and also emerged most frequently in the literature: Recruitment Retention Job satisfaction Absenteeism Turnover Intent to leave/continue practice The incentives were separated into financial and non-financial incentives in addition to incentives for rural and remote areas and incentives for new graduates. Non-financial incentives were categorized as: Positive work environments (workload, autonomy, clinical and social support, and work-life balance); Supports for career and professional development (promotional opportunities, clinical supervision, and educational programs); Work design (staffing models, employment status, and collaboration). Workforce Research and Evaluation, Alberta Health Services 16

20 2.1 Effectiveness of Financial Incentives The most common incentives reported within the reviews revolve around some form of financial compensation. Six reviews examined the relationships between various types of financial incentives (such as pay) and HHR outcomes (Buykx et al., 2010; Chenoweth et al., 2010; Davey et al., 2009; Dolea, Stormont & Braichet, 2010; Hunter & Nicol, 2002; Patterson et al., 2010). Patterson et al. (2010) examined the relationship between financial incentives and job satisfaction and found that higher wages had a positive influence on job satisfaction. Financial incentives may also aid in recruitment and the initial stages of retention. Hunter & Nicol (2002) found evidence that salary was important for the recruitment and retention of occupational therapists. However, other variables are likely involved for ensuring longer-term retention and that a strategy that combines financial compensation with nonfinancial incentives is most effective (Buykx et al., 2010). For example, Dolea, Stormont & Braichet (2010) found evidence that the effectiveness of financial incentives on employee retention declined after five years. Chenoweth et al. (2010) found that financial compensation was not necessarily the highest priority for retaining Australian nurses compared to other factors such as a positive work environment. Davey et al. (2009) reviewed evidence that found some association between pay and absenteeism in American nurses. Some evidence pointed towards a negative relationship while other evidence identified no relationship between the two variables. Finally, Van Ham et al. (2006) found evidence that low income was associated with decreased job satisfaction among general practitioners. 2.2 Effectiveness of Non-Financial Incentives Positive Work Environments Positive work environment includes an array of components such as workload, level of professional autonomy, availability of clinical and social supports, and ability for work-life balance, all potentially affecting HHR outcomes. Autonomy: Professional autonomy has been defined as the extent to which health care providers have the freedom to act on what they know (Aiken, Sloane, & Lake, 1997) and has been found to be an important predictor for job satisfaction and retention. Chenoweth et al. (2010) showed a positive influence (significance not reported) of autonomy on the retention decisions of nurses. Hunter & Nicol (2002) reported clinical autonomy (i.e., participating in decisions) as being important for job satisfaction and retention of occupational therapists. Buykx et al. (2010) reported that the degree of autonomy in the workplace is an important retention strategy. However, in a review by Davey et al. (2009) Workforce Research and Evaluation, Alberta Health Services 17

21 that examined predictors of short-term absences of staff nurses, autonomy was found to not have a significant relationship with absenteeism. Three reviews (Van Ham et al., 2006; Zangaro & Soeken, 2007; Lee & Cummings, 2008) examined the relationship between autonomy and job satisfaction. Lee and Cummings (2008) found no relationship between autonomy and job satisfaction for nurse managers. However, there may have been methodological errors in the study as the findings were contrary to what was expected. On the other hand, in the meta-analysis performed by Zangaro and Soeken (2007), a positive relationship between autonomy and job satisfaction was reported. Similarly, Van Ham et al. (2006) found autonomy to be associated with increased job satisfaction in general practitioners. Clinical and social support: Two reviews investigated the relationship between clinical and social support and job satisfaction (Patterson et al., 2010; Lee & Cummings, 2008). Patterson et al. (2010) linked social support to job satisfaction. This review reported mixed findings for the relationship between social support and job satisfaction and no relationship between supervisory support and job satisfaction. In a review of predictors of job satisfaction for nurse frontline managers, Lee & Cummings (2008) reported higher levels of job satisfaction in settings where managers perceive organizational support and social support from their supervisors. Lee & Cummings (2008) also found that managers who participated in organizational processes were more satisfied with their jobs than those who did not. Workload: Workload and related factors such as job demand, role overload, task complexity, work variety, and work responsibilities have been examined in relation to job satisfaction, retention and absenteeism in four reviews (Patterson et al., 2010; Chenoweth et al., 2010; Davey et al., 2009; Van Ham et al., 2006). Patterson et al. (2010) reviewed these workload variables in relation to job satisfaction and found limited evidence of a relationship. There was no conclusive evidence to support a relationship between job demand and job satisfaction. Finally, Patterson et al. (2010) reported that task complexity had a significant positive impact on job satisfaction. Similarly, routine task had a significantly negative impact on job satisfaction. Van Ham et al. (2006) found that increased workload and work hours were related to decreased job satisfaction, while the amount of variety in the work was related to increased job satisfaction among general practitioners. Workload variables have also been examined in relation to retention and absenteeism. A review by Davey et al. (2009) report mixed findings for work responsibilities and absenteeism. There was some evidence presented reporting significant associations, while other evidence showed no significant associations. Overall, the authors concluded that responsibilities were not associated with absenteeism. In another review, Patterson et al., (2010) reported that high job demands (when in addition to high control) can lead to low absence. Finally, in relation to retention, Chenoweth et al. (2010) concluded there was no Workforce Research and Evaluation, Alberta Health Services 18

22 high quality evidence linking nurse retention with nurse workload, although evidence was reviewed that identified high workload as a direct factor in nurse absenteeism. Work-life balance: There is evidence from three reviews that work-life balance incentives such as a self-scheduling system, flexibility in schedules, family-friendly policies and social hours improve health care provider retention (Butler et al., 2011; Chenoweth et al., 2010; Hunter & Nicol, 2002). Specifically, Butler et al. (2011) reported on the effect of introducing a self-scheduling system on various staff-related outcomes. The self-scheduling system, which allowed staff more input on staffing scheduling on their units, was related to a reduction in turnover rates. Chenoweth et al. (2010) found that flexibility in work schedules, family-friendly policies, social hours and work-life balance all had a positive influence on the decision of nurses to remain in continuing care settings. Hunter & Nicol (2002) reported that flexible working conditions and childcare facilities were important for retention or considering returning to work after having children. Supports for Career and Professional Development Supports for career and professional development include promotional opportunities, clinical supervision and educational programs and have been identified as valued incentives for health care providers (Davey et al, 2009; Chenoweth et al., 2010; Lee & Cummings, 2008; van Wyk & Pillay-Van Wyk, 2010). Promotional opportunities: The review by Davey et al. (2009) examined predictors of shortterm absences of staff nurses and reported on the relationship between promotional opportunities and absenteeism defined as not coming to work when scheduled, and measured by frequency or duration of work days missed. The authors found a negative and significant relationship between promotional opportunities and absenteeism. Clinical supervision: Chenoweth et al. (2010) found that good clinical supervision had a positive influence on recruitment and retention and worked best when there was a good supervisor-mentee relationship. Moreover, clinical supervision had a significant positive effect on nurses perceptions of their role, their feelings of control, and their feelings of being valued by their manager. These factors were particularly important in deciding to remain employed in a study conducted on dementia care wards. Clinical supervision also had a positive influence on retention for nurse managers in continuing care due to improving situational control, coping skills and critical thinking and the development of more future-directed competence. This perceived competence was a key factor in the intention to remain employed in nursing management (Chenoweth et al., 2010). In addition, Hunter & Nicol (2002) found supervision to be an influential factor in the recruitment and retention of occupational therapists. Educational programs: Four reviews investigated the relationship between educational interventions and HHR outcomes (Hunter & Nicol, 2002; Lee & Cummings, 2008; van Wyk Workforce Research and Evaluation, Alberta Health Services 19

23 & Pillay-Van Wyk, 2010; Chenoweth et al., 2010). Lee & Cummings (2008) reported that education and training opportunities increased managers job satisfaction. Van Wyk & Pillay-Van Wyk (2010) found that stress management training had no effect on job satisfaction. Chenoweth et al. (2010) reviewed factors that attract and retain nurses in continuing care and found that educational opportunities including leadership training, a novice to expert residency program, a person-centred care-based curriculum on aged and dementia care, and a three-month training program on dementia aggression have a positive influence on nurse recruitment. Hunter & Nicol (2002) found evidence that training and development, continued education, and professional growth opportunities improve recruitment and retention for occupational therapists. Work Design Work design encompasses staffing models, employment status and nurse-physician collaboration. These incentives have been examined in relation to multiple HHR outcomes described below. Staffing models: Three reviews (Hodgkinson et al., 2011; Patterson et al., 2010; Butler et al., 2011) investigated the relationship between diverse forms of staffing models and specific HHR outcomes including job satisfaction, absenteeism and turnover. Patterson et al. (2010) conceptualize staffing models as re-engineered work design, patient allocation, ward practices, deployment of work and team work. In their review, they found a positive influence between staffing models and job satisfaction. In contrast, Hodgkinson et al. (2011) found that staffing models (primary care model consisting of 24-hour accountability and decision making by one nurse, case method of assignment, direct communication between caregivers, and shift of head nurse role to facilitator) had no influence on job satisfaction or absenteeism. Similarly, a primary nursing model used in the Netherlands (where the primary nurse assesses, prioritises, plans and evaluates the patient s care to coordinate the responsibilities with other nursing staff) did not have a significant effect on absenteeism rates (Butler et al., 2011). However, Patterson et al. (2010) found a positive influence between nurse staffing models and absenteeism. The impact of nursing models on turnover rates is inconclusive. Butler et al. (2011) reported a significant negative relationship between a primary nursing model (described above) and turnover rates. Patterson et al. (2010) reported a positive influence between nurse staffing models and turnover and Hodgkinson et al. (2011) found no relationship Employment status: Chenoweth et al. (2010) and Davey et al. (2009) reviewed various forms of staff skill mix and ratios of full-time to part-time staff in relation to absenteeism and retention. In a review of nurses absenteeism in hospitals, the ratio of part-time staff had a significant negative influence on absenteeism (Davey et al., 2009). In contrast, skill Workforce Research and Evaluation, Alberta Health Services 20

24 mix (defined as the ratio of American graduate nurse-filled full-time positions to baccalaureate of science nurse-filled full-time positions) was not a significant predictor of absenteeism (Davey et al., 2009). Chenoweth et al. (2010) reported on the relationship between staff skill mix and retention and did not find any evidence to suggest that having the right staff skill mix is effective in retaining health care providers. Collaboration: Zangaro and Soeken (2007) in their meta-analysis of studies of nurses job satisfaction found evidence of a positive relationship between nurse-physician collaboration and job satisfaction. Nurse-physician collaboration was conceptualized as jointly sharing information for decision making and problem solving (p.446). Nursephysician collaboration showed moderately positive and significant relationships with job satisfaction in several studies. Finally, Hunter & Nicol (2002) identified multi-professional team work as a main factor of job satisfaction for occupational therapists. Table 3 presents, in matrix form, the various incentives and outcomes on which each review reports. Table 3: Outcomes & Incentives Financial incentives Wages, other financial incentives Staff recruitment Hunter 2002 Non-financial incentives Positive work environments Staff retention Chenoweth 2010 Dolea 2010 Hunter 2002 Autonomy Buykx 2010 Chenoweth 2010 Hunter 2002 Clinical and social support Workload Work-life balance Chenoweth 2010 Chenoweth 2010 Hunter 2002 Supports for career and professional development Promotional opportunities Clinical supervision Educational programs Chenoweth 2010 Hunter 2002 Chenoweth 2010 Chenoweth 2010 Hunter 2002 Staff / Job satisfaction Dolea 2010 Patterson 2010 van Ham 2006 Hunter 2002 Lee 2008 van Ham 2006 Zangaro 2007 Lee 2008 Patterson 2010 Patterson 2010 van Ham 2006 Hunter 2002 Lee 2008 van Wyk 2010 Staff turnover Intent to leave Staff absenteeism Butler 2011 Davey 2009 Davey 2009 Chenoweth 2010 Davey 2009 Patterson 2010 Davey 2009 Workforce Research and Evaluation, Alberta Health Services 21