BMJ Open. For peer review only - Journal: BMJ Open. Manuscript ID: bmjopen

Transcription

1 Patient Satisfaction and Non-U.K. Educated Nurses: A Cross Sectional Observational Study of English National Health Service Hospitals Journal: BMJ Open Manuscript ID: bmjopen-0-00 Article Type: Research Date Submitted by the Author: -Jul-0 Complete List of Authors: Germack, Hayley; University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research Griffiths, Peter; University of Southampton, Faculty of Health Sciences Sloane, Douglas; University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research Rafferty, Anne Marie; King's College London, Florence Nightingale School of Nursing and Midwifery Ball, Jane; National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care (Wessex),University of Southampton, Centre for Innovation and Leadership in Health Sciences Aiken, Linda; University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research <b>primary Subject Heading</b>: Nursing Secondary Subject Heading: Patient-centred medicine, Health services research Keywords: Quality in health care < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, International health services < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, Human resource management < HEALTH SERVICES ADMINISTRATION & MANAGEMENT

2 Page of BMJ Open RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses Patient Satisfaction and Non-U.K. Educated Nurses: A Cross Sectional Observational Study of English National Health Service Hospitals Corresponding author: Linda H Aiken, Center for Health Outcomes and Policy Research, University of Pennsylvania School of Nursing, Curie Blvd, Philadelphia, PA 0- USA. laiken@nursing.upenn.edu Telephone: 0--- Hayley D Germack, PhD, RN Center for Health Outcomes and Policy Research University of Pennsylvania School of Nursing Philadelphia, PA USA Peter Griffiths, PhD National Institute for Health Research Collaboration for Applied Leadership in Health Research Wessex University of Southampton Southampton, England, UK Douglas M Sloane, PhD Center for Health Outcomes and Policy Research University of Pennsylvania School of Nursing Philadelphia, PA USA Anne Marie Rafferty, CBE, FRCN King s College London National Nursing Research Unit Florence Nightingale Faculty of Nursing and Midwifery London, England, UK Jane E Ball National Institute for Health Research Collaboration for Applied Leadership in Health Research Wessex University of Southampton Southampton, England, UK Linda H Aiken, PhD, FAAN, FRCN, RN Center for Health Outcomes and Policy Research University of Pennsylvania School of Nursing i

3 Page of RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses Philadelphia, PA USA Keywords: Quality of healthcare; patient satisfaction; international nurses Word Count: ii

4 Page of BMJ Open RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses iii ABSTRACT Objectives To examine whether patient satisfaction with nursing care in National Health Service (NHS) hospitals in England is associated with the proportion of non-u.k. educated nurses providing care. Design Cross-sectional analysis using data from the 00 NHS Adult Inpatient Survey merged with data from nurse and hospital administrator surveys. Logistic regression models with corrections for clustering were used to determine whether the proportions of non-u.k. educated nurses were significantly related to patient satisfaction before and after taking account of other hospital, nursing, and patient characteristics. Setting English NHS trusts Participants,0 patients years of age and older with at least one overnight stay that completed a satisfaction survey;, bedside care nurses who completed a nurse survey; and NHS trusts Main outcome measure Patient satisfaction Results The percentage of non-u.k. educated nurses providing bedside hospital care, which ranged from % to % of nurses, was significantly associated with patient satisfaction. After controlling for potential confounding factors, each 0 point increase in the percentage of non- U.K. educated nurses diminished the odds of patients reporting good or excellent care by % (OR=0.), and decreased the odds of patients agreeing that they always had confidence and trust in nurses by % (OR=0.). Other indicators of patient satisfaction also revealed lower satisfaction in hospitals with higher percentages of non-u.k. educated nurses.

5 Page of RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses iv Conclusions Use of non-u.k. educated nurses in English NHS hospitals is associated with lower patient satisfaction. Importing nurses from abroad to substitute for domestically educated nurses may negatively impact quality of care. Word Count:

6 Page of BMJ Open RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses v Article Summary Strengths and limitations of this study: This is the first quantitative study to determine the association between employment of nurses trained abroad and patient satisfaction in a systematic national sample of hospitals in England. Unique data previously unavailable enable a rigorous analysis of alternative explanations for lower patient satisfaction associated with high employment of nurses educated abroad. The findings hold important policy implications for workforce planning. The study uses cross-sectional data, and while a number of alternative explanations are considered in our models we cannot rule out the possibility that omitted variables may contribute to associations found.

7 Page of Patient Satisfaction and Non-U.K. Educated Nurses: A Cross Sectional Observational Study of English National Health Service Hospitals INTRODUCTION Maintaining acceptable quality of health care is increasingly challenging under current circumstances of slow economic growth and rising health care expenditures. Two recent high profile reports on quality of care in English National Health Services (NHS) hospitals called attention to egregious lapses in care quality in one hospital[] and persistently high mortality in NHS hospital trusts[]. Both reports made an explicit link between poor quality of care and inadequate nurse resources. Nurses have been caught in the quality storm with some blaming nurses for having uncaring attitudes while researchers produce evidence that the problem is the under-resourcing of nurses in NHS hospitals[, ]. Shortages of nurses at the hospital bedside result both from an inadequate national supply of nurses as well as too few budgeted positions for nurses in hospitals. In the future, England will need to replace large numbers of nurses reaching retirement age. Retirements, coupled with a growing demand for health care by an aging population and increasing prevalence of chronic health conditions create the strong possibility of impending shortages of nurses nationally and locally without policy intervention. Thus the timing is right for a thorough examination of strategies for increasing nurse supply. For decades, the U.K. has resorted to recruiting nurses from other countries when faced with shortages at home[, ]. At the peak of U.K. international nurse recruitment in 00, more than half of new nurses on the U.K. register were educated abroad[]. The U.K was criticized for recruiting nurses from countries suffering from their own nursing shortages, including a highly publicized plea from Nelson Mandela for the NHS to ban recruitment from South Africa[]. Following recent national concern about low hospital nurse staffing levels putting patients at risk, there has been an upturn in hospital recruitment of nurses abroad with NHS trusts reportedly spending at least. million in the last two years recruiting nurses from Portugal, Spain, and Romania in Europe, as well as from the Philippines and India[0]. A recent study examining the association between proportion of non-u.s. educated nurses practicing in U.S. hospitals and 0-day mortality following common surgical procedures found significantly higher risk-adjusted mortality in hospitals that employ higher proportions of non-u.s. educated nurses,

8 Page of BMJ Open after controls for potentially confounding factors[]. Surprisingly little research has been undertaken in the U.K., considering long term dependence upon nurses educated abroad, to determine whether there are quality of care considerations in importing nurses trained in other countries to fill gaps in the U.K. s national nurse supply. There is growing evidence suggesting an association between nursing and patient satisfaction. Nursing has been found to be a major factor associated with patient satisfaction[, ]. The EU-funded RNCAST study found that patients cared for in hospitals with better nurse staffing and good nurse work environments were significantly more likely to rate their hospitals highly, to be willing to recommend their hospitals, and to be satisfied with nursing care[]. The current study explores whether employment of nurses educated outside the U.K. is associated with patient satisfaction in NHS hospitals in England. METHODS Data sources and samples We studied patient reports of satisfaction in NHS trusts in England (the largest of the four nations comprising the U.K. with approximately % of the total population) using data from 00 from three sources: ) the NHS Adult Inpatient Survey[], which asked patients about the overall quality of care they received as well as their perceptions of nurses and other staff; ) the RNCAST England Nurse Survey, which asked nurses (among other things) about their workloads, the work environment in their hospital, and where they received their nursing education[]; and ) the RNCAST Organisational Profile Survey in England that requested information from hospital administrators about selected organizational characteristics of their hospitals, including hospital size, teaching status, and types of available technology[]. Hospital Trusts A random sample of NHS trusts in England stratified by geographic area, size, and teaching status was invited to participate in RNCAST. Thirty-one trusts, representing hospitals, agreed to participate and are included in the study reported here. Hospital administrators from the trusts provided information on the size, teaching status, technology status, and location of the trusts, which we use to describe the different trusts and control for in our analyses of the effects of non-u.k. educated nurses on patient satisfaction. Trusts were categorised as large if they had more than 0 beds, which was the average (mean) number of beds across all trusts in the sample. Technology status was defined as high if the trust had the

9 Page of capacity to perform open-heart surgery or organ transplantation. Teaching status was determined by whether the Trust was affiliated to a medical school and provided training to undergraduate student doctors. Location referred to whether the trust was urban or rural and whether it was inside or outside of London. Other hospital characteristics were measured by aggregating responses from the nurses surveyed, as described below. Nurses All registered nurses on general medical and surgical units in the trusts (up to a maximum 0 units per hospital) were invited to complete a survey[]. Surveys were received from, nurses out of, nurses, a response rate of %. Our nurse sample consists of an average of nurses per trust (with a range of nurses to 0 nurses per trust). Non-U.K. educated nurses were self-identified from the nurse survey as having received their basic professional nursing education in a country other than the U.K. The non-u.k. educated nurses comprised % of respondents. The proportion of non-u.k. nurses varied by trust from % to % of all nurses. Responses from individual nurses were aggregated to create trust-level measures of the nurse work environment and nurse staffing. The work environment was measured using the Practice Environment Scale of the Nursing Work Index (PES-NWI), an extensively validated instrument[-] endorsed by the National Quality Forum[, ]. Nurses were presented with a battery of items on the survey and asked to indicate their level of agreement (using a four-point scale ranging from strongly agree to strongly disagree) that certain organizational features were present in their jobs. Their responses were then used to produce PES-NWI subscales: nurse involvement in hospital affairs, nursing foundations for high-quality patient care, nurse manager leadership, staffing and resource adequacy, and nurse-physician relationships. A summary measure was calculated for each hospital representing the sum of subscales above the median, after individual nurse responses were averaged in each hospital. Hospitals in which nurses rated four or five subscales above the mean were classified as hospitals with better work environments, those with two or three subscales above the median were classified as mixed; and hospitals with none or one subscale above the median were classified as having poor work environments. Nurse staffing was measured by averaging the number of patients that each nurse reported caring for on their last shift[] across all nurses in the same trust. Those who reported

10 Page of BMJ Open caring for greater than patients or less than one were excluded from the sample, assuming they were not providing direct patient care. Lower patient-to-nurse ratios indicate more favorable staffing. Patients Our analysis includes reports from,0 patients receiving care in the participating trusts. Data were obtained from the 00 NHS Adult Inpatient Survey. The questionnaire, administered annually, covers the spectrum of care from arrival at the hospital to discharge. Patients were eligible for the survey if they were years or older, had at least one overnight stay, and were not under the care of a consultant from maternity or psychiatric specialties. The NHS received responses from over,000 patients, a response rate of % of which,0 patients were discharged from hospitals in the trusts included in our study. There were on average 0 patient surveys returned from each sampled trust. Patient satisfaction with care was measured by seven survey items, including an overall patient rating of care, whether the patient wanted to complain about care, whether the patient was treated with respect and dignity, whether explanations about medications were given and understandable, whether nurses answered questions clearly, whether the patient had confidence and trust in nurses providing care, and whether there were enough nurses. Patient characteristics included as control variables in our analyses were patients age, sex, length of stay (LOS), admission type (emergency or planned), and the presence of limiting long term conditions, including deafness or hearing impairment, blindness or partial sightedness, illnesses (i.e., cancer, HIV, diabetes, chronic heart disease, or epilepsy), physical conditions, mental health conditions, and learning disabilities. Analysis strategy We first compared characteristics of U.K. and non-u.k. educated nurses in the sample, using F-tests (for continuous variables) and chi-square tests (for categorical variables) to determine their significance, and provide information on the country in which non-u.k. educated nurses received their education. We examined characteristics of the patients in the sample using means and standard deviations for continuous variables and percentages for categorical variables. We then examined hospital trust characteristics and how they differ across trusts with low (<%), moderate (-0%), and high (>0%) proportions of non-u.k. educated nurses. Chisquare tests were used for categorical variables to evaluate differences between the groups.

11 Page 0 of To estimate the effects of the proportion of non-u.k. educated nurses in the different trusts on patient reports of satisfaction, we used robust logistic regression models with Huber/White sandwich estimators to adjust the standard errors for the clustering of patients within trusts. We first estimated bivariate models, which show the unadjusted main (or direct) effects of a 0% increase in the proportion of non-u.k. educated nurses on each of the patient outcome measures, without additional controls. We then estimated adjusted models which show the main effect of a 0% increase in the proportion of non-u.k. educated nurses after controlling for trust characteristics (size, technology status, nurse staffing, and the quality of the practice environment) and patient characteristics (age, sex, LOS, admission type, ward and the presence of limiting long-term conditions). Analyses were conducted using STATA V.[]. RESULTS Descriptive statistics Demographic characteristics of the nurses in this study are shown in Table, comparing non-u.k. educated nurses (n=) and U.K.-educated nurses (n=). While the average age of the nurses was similar in both groups, at just under 0 years, non-u.k. educated nurses were more likely to be male (% vs. %) and had more years of nursing experience (. ±. years vs.. ± years). The primary countries in which the non-u.k. educated nurses received their nursing education included the Philippines (0%), India (%), and various countries in Africa (%). Some % of the non-u.k. educated nurses were from other European countries. Table. Nurse Characteristics in the Study Hospitals (N=,) Nurse Characteristics Non-U.K. Educated Nurses (N =) U.K. Educated Nurses (N = ) p-value Age (mean ± SD). ±.. ± Years of experience (mean ± SD). ±.. ± <0.00 Male (N and %) Male (%) (%) <0.00 Female (%) (%) Country of education (N and %) Philippines (0%) India (%) Africa (all countries) (%) Europe (non-u.k.) (%) Other Asian (%)

12 Page of BMJ Open Other Western (% Other (%) Missing/invalid (%) P-values are based on F-tests or (in the case of percent male) on the chi-square test. Includes U.S., Canada, Australia, and New Zealand Includes Saudi Arabia, Caribbean countries, and South American countries. The characteristics of the,0 patients in our sample are displayed in Table. The average length of stay (LOS) for these patients was just under days, though the sizable standard deviation associated with the average LOS (over days) indicates that many patients had hospitalizations of much longer duration. More than half of the patients surveyed (%) were over the age of and just over half (%) were women. The majority of patients (0%) were emergency admissions, and the largest numbers of these patients were discharged from general medical wards (%), general surgical wards (%) and trauma and orthopedics wards (%). More than half of the sample reported having at least one limiting long term condition along with the reason for their hospitalization; % reported a single limiting long term condition and % reported multiple conditions. Satisfactory outcomes were expressed to the eleven items indicating satisfaction with care by between % of the patients, who indicated that there were always enough nurses on duty, and % of the patients, who indicated they were always treated with respect and dignity in the hospital and always had trust and confidence in the nurses. Table. Patient Characteristics and Outcomes in the Study Hospitals (N=,0) Patient Characteristics Average length of stay (mean ± SD). days ±. days Age group (N and %) - 00 (%) -0 (%) - 0 (%) + (%) Sex (N and %) Male (%) Female (%) Type of hospitalization (N and %) Emergency (0%) Non-emergency (0%) Discharge Ward General medicine (%)

13 Page of General surgery (%) Trauma & orthopedics (%) Cardiology (%) Urology (%) Gynaecology 0 (%) Geriatric Medicine (%) All other (%) Limiting Long Term Conditions(N and %) None (%) One (%) Two or more 0 (%) Outcomes (Satisfaction) Rated care received as very good or excellent 0 (%) Didn t want to complain about care. 0 (%) Always treated with respect and dignity while in the hospital 0 (%) A member of staff always explained the purpose of medicines. (%) Nurses always provide easy to understand answers 0 (%) Always have confidence and trust in nurses 0 (%) There were always enough nurses on duty to care for the patient in the hospital (%) Limiting long term conditions include deafness or hearing impairment, blindness or partial sightedness, illnesses (i.e., cancer, HIV, diabetes, chronic heart disease, or epilepsy), physical conditions, mental health conditions, and learning disabilities. Question only asked of patients receiving medications and who indicated they sometimes needed medications explained. Note: Data were missing for no more than % or respondents on any item. Table displays the characteristics of the trusts, overall and by proportion of non- U.K. educated nurses. Overall, roughly one-third of the trusts were high technology, and roughly in 0 were teaching hospitals. Twenty-six of the hospital trusts were located outside of London, one-third of them had workloads involving more than eight patients per nurse, and more than one-third had poor work environments. Compared to trusts with low (n=) or moderate (n=) proportions of non-u.k. educated nurses, those with a high proportion (n=) were significantly more likely to be located in London and have more than 0 beds. Trusts with different proportions of non-u.k. educated nurses did not differ significantly with respect to technology, teaching status, nurse staffing, or quality of the nurse work environment. Table. Characteristics of the Hospital Trusts (N = ), Overall and by the Proportion of Non-U.K. Educated Nurses (FENs)

14 Page of BMJ Open Characteristic (N and %) All Trusts (N = ) Hospital Trust Group Trusts with - High Middle Low Proportion Proportion Proportion of FENs of FENs of FENs (N = ) (N = ) (N = ) p-value Technology High (%) (%) (%) (%) 0.0 Low 0 (%) (%) (%) (%) Hospital size Under 0 beds (%) (%) (%) (%) beds or more (%) 0 (%) (%) (%) Teaching Status Teaching hospital (%) (%) (%) (%) 0. Non-teaching hospital (%) (%) (%) (%) Location London (%) (%) 0 (0%) 0 (0%) 0.00 Not London (%) (%) (00%) (00%) Staffing < patients per nurse (%) (%) (%) 0 (0%) patients per nurse (%) (%) (%) (%) > patients per nurse 0 (%) (%) (%) (%) Work environment Poor (%) (%) (%) (%) 0.0 Mixed (%) (%) (%) (%) Better (%) (%) (%) (%) P-values are based on chi-square tests. The abbreviation FEN (for foreign educated nurse) is used to denote non-u.k. educated nurses. Hospital trust groups were defined by the percentage of the RN nurses that were not U.K. educated, as follows: High =0%; Middle = -0%; Low <%. Outcomes The estimated effects of having greater proportions of non-u.k. educated nurses on patient satisfaction are shown in Table. The unadjusted odds ratios suggest that before taking account of other factors, a greater proportion of non-u.k. educated nurses had a significant and negative effect on whether patients rated their care as very good or excellent (OR = 0.), and a similarly significant and negative effect on three of the other six indicators of patient satisfaction. After taking account, in the adjusted models, of differences across the hospital trusts in both other hospital characteristics and patient characteristics, the effect of the proportion of non-u.k. educated nurses is even more pronounced, of roughly similar size for all seven outcome measures (odds ratios range from 0. to 0.), and statistically significant for six of the seven.

15 Page of These odds ratios imply, for example, that patients in a hospital with 0% non-u.k. educated nurses would be less likely by a factor of 0. (or % less likely) than patients in hospitals with no non-u.k. educated nurses to rate their care as very good or excellent. Patients in a hospital with 0% non-u.k. educated nurses would be less likely by a factor of 0. = 0. (or % less likely) than patients in hospitals with no non-u.k. educated nurses to rate their care as very good or excellent. Table. Association between the Proportion of Non-U.K. Educated Nurses in Hospital Trusts and Patient Satisfaction Odds Ratios Outcome Unadjusted Adjusted Rate care received as very good or excellent 0.* 0.*** (0. 0.) (0. 0.) Didn t want to complain about care (0..0) (0..0) Always treated with respect and dignity 0. 0.*** while in the hospital (0..00) (0. 0.) A member of staff always explained the 0.* 0.0*** purpose of medicines. (0. 0.) (0. 0.) Nurses always provide easy to understand 0.*** 0.*** answers (0. 0.) (0. 0.0) Always have confidence and trust in nurses 0.*** 0.*** There were always enough nurses on duty to care for the patient in the hospital (0. 0.) (0. 0.) 0. 0.* (0..0) (0. 0.) Odds Ratios refer to the change in the odds of the different outcomes associated with each 0% increase in the proportion of non-u.k. educated nurses in the hospital trusts. Single, double and triple asterisks denote odds ratios that are significant at the 0.0, 0.0, and 0.00 levels, respectively. The adjusted model included controls for patient characteristics (gender, age, limiting long term condition, type of admission [emergency or planned], length of stay, discharge ward) and hospital/trust characteristics (size, technology status, nurse staffing (day nurse), and the practice environment). DISCUSSION Summary of main results This study provides the first empirical evidence that employment of high proportions of non- U.K. educated nurses by NHS hospitals is associated with lower patient satisfaction with care. We found that even after taking account of differences in trusts in nurse staffing, the quality of the nurse work environment, and other features of the trusts including location, every 0%

16 Page of BMJ Open increase in the proportion of non-u.k. educated nurses was associated with a % decrease in the likelihood of patients rating the hospital good or excellent (as opposed to fair or poor) and a similar decrease in the likelihood of patients always having confidence and trust in nurses. The lower patient satisfaction reported by patients cared for in hospitals that employed substantial proportions of non-u.k. educated nurses could not be explained by other features of these hospitals. For example, lower satisfaction was not explained by poorer nurse staffing or poorer work environments of hospitals that employed substantial numbers of non-uk nurses. Neither was the lower observed lower patient satisfaction in hospitals employing substantial numbers of non-u.k. educated nurses explained by the location of hospitals or other features of hospitals such as size, high technology, or teaching status, or by the characteristics of the patients providing ratings. Our findings suggest that there is an important link between substantial numbers of nurses trained abroad and lower patient satisfaction. These findings have important implications for national nurse workforce planning, especially in the current context where there are substantially more U.K. applicants to nursing schools than funded student positions[]. The long-standing U.K. policy of turning to international nurse recruitment rather than investing adequately in the domestic supply of nurses may have negative consequences for quality of care while at the same time limiting opportunities for U.K. citizens to become nurses and thus benefit from the availability of good jobs. Discussion of differences in outcomes The link between lower patient satisfaction and nursing education outside the U.K. is not well understood, the exact explanations are beyond the scope of this study. Language and cultural differences may play a role in the association between higher proportions of non-u.k. nurses and lower patient satisfaction. The majority of hospitalized patients surveyed were years of age or older, and acutely ill older patients may have hearing and communications challenges. Also, transcultural research suggests that professional nursing practice may be influenced by professional norms in the country of origin that are different from professional nursing expectations in an adopted country in ways that could influence quality of care and patient satisfaction, including nurse physician communication[, ]. Indeed, other findings from the RNCAST study show that nurses educated in developing countries now practicing in European hospitals are significantly more likely than European-educated nurses to invest time in

17 Page of performing tasks that may not require the knowledge and skill of a professional nurse, such as cleaning patient rooms and equipment[], while needed nursing care is left undone because of lack of time. The kinds of needed professional nursing care most frequently left undone in hospitals include spending time talking with patients and their families, and educating patients about their self-care after discharge, activities that might be particularly salient to patients and influence their perceptions of their overall hospital experience[]. In the U.K., nurses trained outside the European Union wishing to register to practice are required to undertake a targeted training program, which includes a period of supervised practice for a minimum of three months, as well as 0 days of education relevant to the practice of nursing in the U.K. covering the structure of the U.K. health care system, the Code of Professional Conduct, and the accountabilities of a registered nurse. Perhaps the overseas nursing programme, designed to ensure basic competencies and fitness to practice, is not sufficient to ensure high quality patient-centered care. Variability in the quality of adaptation programmes in the U.K. has been noted in the past[]. More extensive preparation of foreign educated nurses may be required. If this is the case, the expediency of recruiting nurses from other countries to fill short term supply deficits may be lost. This, combined with the deleterious effects of overseas recruitment on healthcare quality and access in many of the exporting countries, emphasizes the importance of planning for and investing in an adequate supply of domestically educated nurses. Large hospitals and those in London are particularly likely to employ larger proportions of non-u.k educated nurses. In many cases, these are hospitals known for high quality of care and in general have good resources; thus, it is even more notable that despite good resources, patient satisfaction is lower than expected, seemingly related to high proportions of non-u.k. educated nurses. In addition to a review of the required training program for international nurses, at least two other types of policy interventions may be warranted. The number of student places commissioned in London has declined more than in other parts of the country[], calling into question the effectiveness of devolving decision-making on nurse supply to the local trust level. Additionally, the very high cost of reasonable quality housing in London may discourage U.K. educated nurses from working in London hospitals[0]. Trusts have the flexibility to adjust nationally determined nurse wage rates to account for high local costs of living but there is little

18 Page of BMJ Open evidence that trusts fully exploit this opportunity to improve recruitment of U.K. educated nurses, potentially because of constrained budgets. Limitations The study uses cross-sectional data which cannot establish causality. Despite patient level risk adjustment and use of multiple trust characteristics to control for potential confounds, we cannot rule out the possibility that variables omitted from our models may be responsible for the associations found. While our study is limited to of the 0 acute NHS trusts in England, participating trusts were obtained through a stratified, random sampling procedure, and the participating hospitals as a group are not significantly different in characteristics from other trusts of more than 00 beds nationally. Moreover our results are robust even without controls for institutional characteristics suggesting that unique features of the participating hospitals are not responsible for the observed association between higher proportion of non-uk educated nurses and lower patient satisfaction. Our data are from 00, which is the latest data available on the variables studied. Also, there is little reason to anticipate that the relationship between non-u.k. nurses and patient satisfaction has fundamentally changed in the intervening years. CONCLUSIONS Utilization of a substantial proportion of non-u.k. educated nurses in English NHS hospitals is associated with lower patient satisfaction with the overall hospital experience, and with lower satisfaction with nursing care specifically. Recent estimates show that one in ten nurses in the U.K. is from another country[0]. Diversity in the nurse workforce is an important goal, but one that could be achieved by making nursing education more accessible to U.K. citizens. Our findings suggest that the use of nurses educated in other countries to substitute for professional nurses educated at home is not without risks to quality of care.

19 Page of Contributors HDG wrote the analysis plan, cleaned and analysed the data, and drafted and revised the paper. LHA supported the interpretation of results, drafted, and revised the paper. DMS, PG, AMR, and JEB supported the interpretation of results and revised the paper. Competing interest All authors declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous years; no other relationships or activities that could appear to have influenced the submitted work Funding This research was funded by the European Union s Seventh Framework Programme (FP/00-0, grant agreement no, W Sermeus, PI) and the Rita & Alex Hillman Foundation Scholars in Nursing Innovation. Data Sharing All authors had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis.

20 Page of BMJ Open REFERENCES. Francis R. Report of the Mid Staffordshire NHS Foundation Trust Public Inquiry. London: The Stationary Office; 0 Feb; Available from: Keogh K. Review into the quality of care and treatment provided by hospital trusts in England: overview report. National Health Service; 0 Jul; Available from: Aiken L, Rafferty A, Sermeus W. Caring nurses hit by a quality storm. Nurs Stand 0;():-.. Aiken L, Sloane DM, Bruyneel L, et al. Nurse staffing and education and hospital mortality in nine European countries: a retrospective observational study. The Lancet 0;():-0.. Buchan J. International recruitment of nurses: policy and practice in the United Kingdom. Health Serv Res 00;( Pt ):-.. Attree M, Flinkman M, Howley B, et al. A review of nursing workforce policies in five European countries: Denmark, Finland, Ireland, Portugal and United Kingdom/England. J Nurs Manag 0;():-0.. Buchan J, Seccombe I, O'May F. Safe staffing levels--a national imperative. The UK nursing labour market review 0. London Royal College of Nursing 0 Sep.. Kingma M. Nurses on the move: migration and the global health care economy. Ithaca, N.Y.: IRL Press; 00.. Buchan J, O'May F, Dussault G. Nursing Workforce Policy and the Economic Crisis: A Global Overview. J Nurs Scholar 0: Keogh K. Overseas recruitment big business again as safe staffing recognised. Nurs Stand. 0 Mar.. Buchan J, Seccombe I. From Boom to Bust? The UK Nursing Labour Market Review 00/. London: Royal College of Nursing, 00.. Kutney-Lee A, McHugh MD, Sloane DM, et al. Nursing: a key to patient satisfaction. Health Aff (Millwood) 00;():w-.. Jha A, Orav E, Zheng J, et al. Patients' Perception of Hospital Care in the United States. New Engl J Med 00;():-.. Aiken LH, Sermeus W, Van den Heede K, et al. Patient safety, satisfaction, and quality of hospital care: cross sectional surveys of nurses and patients in countries in Europe and the United States. BMJ 0;:e.. Neff DF, Cimiotti J, Sloane DM, et al. Utilization of non-us educated nurses in US hospitals: implications for hospital mortality. Int J Qual Health Care 0:.

21 Page 0 of Sermeus W, Aiken LH, Van den Heede K, et al. Nurse forecasting in Europe (RNCAST): Rationale, design and methodology. BMC Nurs 0;0:.. Picker Institute Europe. Guidance Manual for the NHS Adult Inpatient Survey 00. Oxford: The Coordination Centre for the Acute Survey Programme00.. Lake ET. Development of the practice environment scale of the Nursing Work Index. Res Nurs Health 00;():-.. Lake ET. The nursing practice environment: measurement and evidence. Med Care Res Rev 00;( Suppl):0S-S. 0. Aiken LH, Cimiotti J, Sloane DM, et al. Effects of nurse staffing and nurse education on patient deaths in hospitals with different work environments. Med Care 0;():0-0.. Warshawsky NE, Havens DS. Global use of the Practice Environment Scale of the Nursing Work Index. Nurs Res 0;0():-.. National Quality Forum. Practice Environment Scale - Nursing Work Index (PES-NWI) (composite and five subscales). 0 [May, 0]; Available from: National Quality Forum. National voluntary consensus standards for nursing-sensitive care: An initial performance measure set--a consensus report. Washington, DC00; Available from: ursing-sensitive_care An_Initial_Performance_Measure_Set.aspx.. Aiken LH, Clarke SP, Sloane DM, et al. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA 00;():-.. StataCorp. Stata/IC. for Windows College Station, TX0.. Flynn L, Aiken LH. Does international nurse recruitment influence practice values in U.S. hospitals? J Nurs Scholar 00;():-.. Nichols B, Campbell J. The experiences of internationally recruited nurses in the UK (-00): an integrative review. J Clin Nurs 00;(-0):-.. Bruyneel L, Li B, Aiken L, et al. A multi-country perspective on nurses' tasks below their skill level: Reports from domestically trained nurses and foreign trained nurses from developing countries. Int J Nurs Stud 0;0():0-.. Ball J, Murrells T, Rafferty A, et al. Care left undone during nursing shifts: associations with workload and perceived quality of care. BMJ Quality & Safety 0;():-. 0. Osborne H. House prices: eight in 0 homes out of reach of families on average wages. The Guardian. 0 Jun.

22 Page of BMJ Open STROBE Statement Checklist of items that should be included in reports of cross-sectional studies Item No Recommendation Title and abstract (a) Indicate the study s design with a commonly used term in the title or the abstract Introduction The cross-sectional study design is noted in the title and in the Design section of the Abstract. (b) Provide in the abstract an informative and balanced summary of what was done and what was found The study design, settings, participants, outcome measure and results are described in the Abstract. Background/rationale Explain the scientific background and rationale for the investigation being reported The background and rationale are explained in the Introduction (pp. -) Objectives State specific objectives, including any prespecified hypotheses Methods The primary objective of the study is stated in the last sentence of the introduction (p. ) Study design Present key elements of study design early in the paper. Key elements of the study design are described in brief in the Design section of the Abstract and in greater detail in the Methods section of the paper (pp. -) Setting Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection. The setting, locations and dates of the surveys of trusts, patients and nurses are given in the first paragraph of the Methods section (pp.-). Participants (a) Give the eligibility criteria, and the sources and methods of selection of participants. We describe the sampling of trusts, patients and nurses in the Methods section (pp. -). We also reference prior publications which contain additional details (Sermeus et al. 0). Variables Clearly define all outcomes, exposures, predictors, potential confounders, and effect Data sources/ measurement modifiers. Give diagnostic criteria, if applicable. Key patient variables (including satisfaction and control variables) are described in general on pp. -, and additional details on patient variables used for risk adjustment are shown in Table. Characteristics of the trusts sampled and nurses surveyed are described on pp. -, and additional details are given in Tables and. For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group. Data are from patient satisfaction surveys, organisational trust surveys, and nurse surveys, as described on pp. -. Additional information on all measures is provided in Tables - and related text in the results section, on pp. -. Bias Describe any efforts to address potential sources of bias. We attempted to obtain unbiased results by undertaking proper risk adjustment and by using logistic regression models with robust estimation procedures to take account of patients being nested within trusts (see p. ; section.). Study size 0 Explain how the study size was arrived at. We describe the sampling of trusts, patients and nurses in the Data and Methods section (pp. -). We also reference

23 Page of prior publications which contain additional details on the broader RNCAST study from which these data were derived (Sermeus et al. 0). Quantitative variables Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen and why. Some quantitative variables (i.e. staffing, proportion of FENs) were treated as continuous and not grouped. We describe grouping of work environment in the Methods section (pp. -). Statistical methods (a) Describe all statistical methods, including those used to control for confounding We discuss the logistic regression models with robust estimation procedures we used and the control variables we employed in the Analysis Strategy sub-section of the Methods section (p. ). (b) Describe any methods used to examine subgroups and interactions. Preliminary analyses revealed no significant interactions, so results reported are restricted to main effects models. (c) Explain how missing data were addressed We note that our analyses were restricted to patients for whom we had complete data on comorbidities present on admission, surgery type, status at discharge, and other characteristics used for risk adjustment. In table we note that data were missing on those characteristics for less than % of all patients. (d) If applicable, describe analytical methods taking account of sampling strategy. We note (p. ) that our approach took account of the fact that patients were nested within trusts. (e) Describe any sensitivity analyses On p. we describe having estimated the effect of interest (the percent of non-u.k. educated nurses) under alternative specifications. Additional measures of nurse staffing and the work environment were used in alternative models (not reported), with no discernible effect on the results. Results Participants (a) Report numbers of individuals at each stage of study eg numbers potentially eligible, examined for eligibility, confirmed eligible, included in the study, completing follow-up, and analysed. We report that of the sampled trusts agreed to participate (p. ). We also report the response rates for nurses and patients. (b) Give reasons for non-participation at each stage Unknown (c) Consider use of a flow diagram. Not helpful here. Descriptive data (a) Give characteristics of study participants (eg demographic, clinical, social) and information on exposures and potential confounders. Numbers and characteristics of trusts and patients are shown in Tables - (Also see related text). More detailed information on the trusts, patients and nurses in the study can be found in related papers that are cited. (b) Indicate number of participants with missing data for each variable of interest. In table we note that data were missing on those characteristics for less than % of all patients. Outcome data * Report numbers of outcome events or summary measures. The patient satisfaction outcome measures are shown at the bottom of Table. Main results (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their precision (eg, % confidence interval). Make clear which confounders were adjusted for and why they were included. Confidence intervals are given with estimates in Table.

24 Page of BMJ Open (b) Report category boundaries when continuous variables were categorized. The categorization of the nurse work environment is described on pp. - of Methods. (c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period. Not relevant. Other analyses Report other analyses done eg analyses of subgroups and interactions, and sensitivity analyses. All analyses have been reported. Discussion Key results Summarise key results with reference to study objectives (see Discussion and Conclusions pp. -) Limitations Discuss limitations of the study, taking into account sources of potential bias or imprecision. Discuss both direction and magnitude of any potential bias See Limitations, p.. Interpretation 0 Give a cautious overall interpretation of results considering objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence See Discussion and Conclusions sections pp. -. Generalisability Discuss the generalisability (external validity) of the study results. While point estimates of factors (like the nursing characteristics) and satisfaction likely vary from trust to trust, we believe the associations found are likely generalizable to all NHS trusts, as we note in the last paragraph of the discussion section. Other information Funding Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based. (See below title page, p.ii. *Give information separately for exposed and unexposed groups. Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at Annals of Internal Medicine at and Epidemiology at Information on the STROBE Initiative is available at

25 Patient Satisfaction and Non-U.K. Educated Nurses: A Cross Sectional Observational Study of English National Health Service Hospitals Journal: BMJ Open Manuscript ID bmjopen-0-00.r Article Type: Research Date Submitted by the Author: -Oct-0 Complete List of Authors: Germack, Hayley; University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research Griffiths, Peter; University of Southampton, Faculty of Health Sciences Sloane, Douglas; University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research Rafferty, Anne Marie; King's College London, Florence Nightingale School of Nursing and Midwifery Ball, Jane; National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care (Wessex),University of Southampton, Centre for Innovation and Leadership in Health Sciences Aiken, Linda; University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research <b>primary Subject Heading</b>: Nursing Secondary Subject Heading: Patient-centred medicine, Health services research Keywords: Quality in health care < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, International health services < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, Human resource management < HEALTH SERVICES ADMINISTRATION & MANAGEMENT

26 Page of BMJ Open RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses Patient Satisfaction and Non-U.K. Educated Nurses: A Cross Sectional Observational Study of English National Health Service Hospitals Corresponding author: Linda H Aiken, Center for Health Outcomes and Policy Research, University of Pennsylvania School of Nursing, Curie Blvd, Philadelphia, PA 0- USA. laiken@nursing.upenn.edu Telephone: 0--- Hayley D Germack, PhD, RN Center for Health Outcomes and Policy Research University of Pennsylvania School of Nursing Philadelphia, PA USA Peter Griffiths, PhD National Institute for Health Research Collaboration for Applied Leadership in Health Research Wessex University of Southampton Southampton, England, UK Douglas M Sloane, PhD Center for Health Outcomes and Policy Research University of Pennsylvania School of Nursing Philadelphia, PA USA Anne Marie Rafferty, CBE, FRCN King s College London National Nursing Research Unit Florence Nightingale Faculty of Nursing and Midwifery London, England, UK Jane E Ball National Institute for Health Research Collaboration for Applied Leadership in Health Research Wessex University of Southampton Southampton, England, UK Linda H Aiken, PhD, RN, FAAN, FRCN Center for Health Outcomes and Policy Research University of Pennsylvania School of Nursing i

27 Page of RUNNING HEAD: Patient Satisfaction and Non-U.K. Educated Nurses Philadelphia, PA USA Keywords: Quality of healthcare; patient satisfaction; international nurses Word Count: ii