Table 2.

Research evidence around mixed ED clinical staff perceptions of their working environment

Author, year, country	Aim/s	Sample	Research design/tools/ analysis type†	Rigour, reliability, validity‡	Findings	Strengths	Limitations§	Recommendations/ implications	MMAT %
1. Joe, Kennedy and Bensberg, 2002, Australia	To demonstrate a comprehensive workplace health survey that is able to identify indicators that contribute to staff workplace welfare	n = 323/500 staff from seven Melbourne suburban public hospital EDs with similar attendance numbers, case mix and demography n = 59 doctors, n = 198 nurses, n = 30 clerical/admin staff and n = 22 other staff	– §§Quantitative descriptive study using a cross‐sectional correlational design from a self‐reported validated employee survey designed by Service Management Australia (a subsidiary of Marketshare) with terminology within the survey altered to make it relevant to the ED – Included closed, rating (5‐point scale) and open questions for quant and qual content; mixed methods study – Calculated a ‘performance gap’ around key issues – the difference between importance rating and perceived performance rating	– 64% response rate – Employee survey commonly used (Marketshare); refined and validated regularly by research conducted locally and compared with data from abroad	– Staff rated a safe environment, professional standards, and staff morale the most important factors for workplace health. They were most satisfied with the flexibility of work arrangements (86%) and leadership (80%), and were least satisfied with the performance management of staff (69%) and job satisfaction and morale (67%) – The largest gaps between perceived importance and performance were in the provision of safe well‐lit parking, staff morale, and the use of reward and recognition systems	– Utilised widely used mixed method survey tools across a number of sites and a wide range of staff – Explored a range of ED aspects from communication and staff morale to staff injuries	– ††Skewed population; 75% women and 61% nurses – No follow up of non‐respondents to the survey	– Provides direction for further research into ED workplace health, enabling refinement of indicators reflecting various aspects of workplace health, and correlation of indicators with sick leave, stress and injury – Also indicators of how various indicators affect different staff groups and Workplace health in EDs	100
2. McFarlane, Duff and Bailey, 2004, Jamaica, West Indies	To explore factors associated with occupational stress in ED staff and the coping strategies used	28/33 of health personnel working in the A&E n = 15 doctors, n = 8 registered nurses and n = 5 enrolled assistant nurses – Single site	– §§Quantitative descriptive cross‐sectional design using two self‐reported, trialled, quantitative and open‐ended (qualitative) items that included limited demographic information – Open‐ended data were analysed thematically	– Response rate = 85%, 54% doctors, 29% registered nurses and 18% enrolled assistant nurses	– A&E was reported to be stressful, with the major sources of stress reported as the external environment and the amount and quality of the workload and resulted in emotional, physical and behavioural symptoms – Effective use of humour, teamwork and ‘extracurricular’ activities in buffered the effects of stress	– Little published information from West Indian hospitals	– Abstract only available – No evidence of ethical approval	– Increased monetary compensation, more staff and positive feedback from managers as factors that may relieve work stress – Organised counselling and stress management programmes may be useful	25
3. Escriba‐Aguir and Perez‐Hoyos, 2007, Spain	To determine if psychosocial WE differentially altered psychological well‐being for ED clinical staff	SSEM members including ED doctors and nurses Reported n = 639; data collected from n = 278 nurses and n = 358 doctors	– §§Quantitative descriptive study using a cross‐sectional correlational design from a self‐reported validated quantitative questionnaire – Mental health and vitality dimension of the SF‐36 Health survey, emotional exhaustion dimension of Maslach's burn‐out inventor and the job content questionnaire – Descriptive statistics and logistic regression	– Supported by Karasek and Theorell's demand‐control WE model – Careful consideration of potential confounding factors including socio‐professional and gender‐role related variable – Response rate 68%	– Psychosocial WE factors strongly influenced clinical staff psychological well‐being, but the effect varied in nurses and doctors – Doctors were more likely to show low vitality, poor mental health and high levels of emotional exhaustion from high psychological demands – Low levels of job control and co‐workers social support also increased the risk of poor mental health in doctors and the risk of high emotional exhaustion in nurses – Little impact of physical workload on reported well‐being – Lack of supervisor support for doctors but not nurses	– Baseline data comparison with normative data from American health professionals – Comparison across health professionals drawn from a similar clinical pool	– ‡‡, ††Sample all drawn from one professional society – No information provided about the number of EDs or the size/busyness of the ED on results – Limited follow up of non‐respondents to the survey	– Greater need for capacity of control for doctors in EDs – Need for further investigation including the role of professional career choices and work‐family roles on clinical staff in ED – Need to establish improvements in psychosocial WE to reduce the risk of psychological distress in ED clinical staff, especially doctors	75
4. Magid, Sullivan, Cleary et al., 2009, USA	To assess the degree to which ED staff felt that EDs are designed, managed, and supported in ways that ensure patient safety, including the physical work environment, staffing, equipment, supplies, teamwork and coordination with other services	n = 3562 from 69 sites – Participation invited from sites affiliated with the Emergency Medicine Network, postings on emergency medicine list‐servers and through presentations at emergency medicine meetings	– §§Quantitative descriptive cross‐sectional design using a self‐reported, extensively validated 50 question quantitative questionnaire that included limited (5) demographic questions – Multisite research included collecting information about size, complexity of each research site – Response rate 66% random sample of 80 eligible staff at each ED who worked an average of one or more ED clinical shifts per week – Eligible survey respondents included physicians, nurses, nurse practitioners, physician assistants and nursing assistants – Respondents had the option of completing a paper‐based instrument or completing the survey online – Analyses were weighted to adjust for the differential sampling rates, differential nonresponse rates, and respondent job class	– The developed scales generally had good reliability (Cronbach's): physical environment (0.60), staffing (0.65), equipment and supplies (0.93), nursing (0.90), teamwork (0.60), culture (0.79), triage and monitoring (0.91), information coordination and consultation (0.64), and inpatient coordination (0.88) – 69/102 sites initially interested actually took part (68%) – Nonrespondents received two additional surveys at 2‐week intervals, for a total of three surveys during 6 weeks – Adjustment to acceptable alpha level (0.003) due to multi‐testing	– Survey respondents commonly reported problems in four systems critical to ED safety: physical environment, staffing, inpatient coordination and information coordination – Generally, factors around working environment, including ‘blame’ culture, staff supervision, cross‐discipline team work, were rated very highly.	– Multiple step survey development, validation, piloting and testing establishing face and construct validity – Piloted at 10 EDs and then administered to 65 different EDs across the USA	– ‡‡Excluded military, Veterans Administration, and children's hospitals, as well as hospitals in US territories – A modest honorarium was given to survey respondents to minimise nonresponse bias – Four EDs with response rates of 45% or less and individual questionnaires with answers to less than 80% of the survey items were excluded – Results of this study may not be generalisable to all EDs, because the facilities that participated in this study tended to be larger, were more urban and were more likely to be affiliated with an emergency medicine training programme than the typical US ED – Self‐selection bias; namely, the institutions that participated in the study were more interested in and more aware of concerns surrounding patients safety than the average US ED	– Substantial improvements in institutional design, management, and support for emergency care are necessary to maximise patient safety in US EDs	100
5. Healy and Tyrrell, 2011, Ireland	To examine nurses' and doctors' attitudes to, and experiences of, workplace stress in three EDs	n = 103/150 – Convenience sample of n = 90 nurses and n = 13 doctors from three EDs	– Descriptive cross‐sectional design from a self‐reported 16 item survey with a mixture of yes/no, Likert‐type (quantitative) item response and open‐ended questions and some additional experience and demographic data – Descriptive statistics (count, mean ± SEM) and χ2/Mann–Whitney U comparative tests	– 69% response rate – Non‐validated, unpiloted, author developed survey may compromise survey validity	– Most commonly identified stressors were ‘work environment’ including rostering, workload, crowding, traumatic events, shift work, doctor turnover, inter‐staff conflict, poor teamwork and poor managerial skills – Generally, older and more experienced staff reported less stress – Little managerial/workplace support to manage stress – Violence in the workplace and death/resuscitation of critically ill patients were frequently cited as stressful	– Open questions enabled ED staff to construct their own descriptions of stressors and stress priorities, creating a rich dataset	– No information about survey follow up, preservation of anonymity, hospital size provided– 10% nurses and 61% doctors were men – No multifactorial analysis	– ED staff need protect from relentless stress, particularly younger less experienced staff who are most vulnerable to the effects of stress – Managers must establish a supportive culture that recognises the real issues around staff stress and demonstrate they value staff	50
6. Flowerdew, Brown, Russ, Vincent and Woloshynowych, 2012, London, UK	To identify key stressors for ED staff, explore positive and negative behaviours associated with working under pressure and consider interventions that may improve ED team functions	Purposive sampling recruitment of medical and nursing staff of varying seniority – Recruitment continued until no significant new themes emerged during interview (‘theoretical saturation’) – n = 22 staff, n = consultants, n = 7 registrars, n = 5 lower grade doctors, n = 6 nurses – Single site	– §§Qualitative – Semi‐structured interviews were recorded and anonymously transcribed and analysed to extract broad themes from the interviews, and responses were coded using the NVivo computer programme	– Themes were independently confirmed by a second researcher – Coded material was subject to member check to reduce investigator bias	– Identified stressors included the ‘4 h’ targets, excess workload, staff shortages and lack of teamwork, both within the ED and with inpatient staff – Leadership and teamwork are mediating factors between objective stress (e.g. workload and staffing) and the subjective experience – Impacts of high pressure on communication practices, departmental overview and the management of staff and patients, as well as high levels of misunderstanding between senior and junior staff – Effective leadership and teamwork training, staff breaks, helping staff to remain calm under pressure and addressing team motivation all part of the solution	– Information drawn from a variety of clinical staff using an open‐ended set of questions to allow themes to emerge – Study explored problems and possible solutions – Use of direct quotes adds participants' ‘voices’	– ††Interviews reflect self‐reported behaviour and may be biassed by ineffective recollection, misunderstanding or embarrassment	– Identified that many ED staff lack training in coping strategies and in ‘non‐technical skills’ such as communication, situational awareness and leadership that could be rectified – Building a resilient team with strong leadership is integral to being able to withstand the pressures of the ED	100
7. Yates, Benson, Harris and Baron, 2012, UK	To compare levels of psychological health in medical, nursing and administrative staff from a UK ED with a comparative orthopaedic department – Also, to investigate the influence of coping strategies and the support people receive from their colleagues (i.e. social support)	n = 136 – Emergency (n = 73) and orthopaedic (n = 63) staff – Single site (two departments)	– §§Quantitative cross‐sectional design using four self‐reported quantitative surveys including the General Health Questionnaire‐12 (GHQ12), the Hospital Anxiety and Depression Scale (HADS), the Brief COPE consisting of 14 scales, each of two items and a brief (three‐item) measure of social support – Descriptive statistics (percentage and correlation)	– Insufficient information provided to comment – 73 (50%) ED staff (30 nurses, 19 doctors, 24 administrative staff) and 63 (39%) OD staff (32 nurses, 16 doctors, 15 administrative staff)	– Proportion of staff experiencing clinically significant levels of distress was higher than would be expected in the general population – Better psychological health was associated with greater use of problem‐focused coping and less use of maladaptive coping – Social support was associated with better psychological health and greater use of problem‐focussed coping – Emergency physicians, but not other ED staff, reported an increased risk of psychological distress – Increased psychological health was associated with the use of problem‐focused coping strategies and higher levels of social support at work – Those reporting lower levels of psychological health were more likely to use maladaptive coping strategies	– One of the few studies incorporating a direct control/comparison group	– ††, ‡‡Little information supplied regarding recruitment and participation – No ethical approval noted – Overall sample was heterogeneous (ED/OD) and the sub‐samples were relatively small and did not allow an analysis of potential differences between staff groups within each department – Reliance on self‐reported survey data may have compromised the richness of data	– Priority should be given to developing and evaluating interventions to improve psychological health in ED staff – Coping strategies and social support are important factors to incorporate into such interventions to improve psychological health in ED staff – Use of problem‐focused coping strategies and social support may be important factors to incorporate into intervention – Need for further research	25
8. Ajeigbe, McNeese‐Smith, Leach and Phillips, 2013, USA	To examine the impact of a teamwork training protocol on perception of job environment, autonomy and control over practice in ED clinical staff	n (intervention) = 166 RNs and 25 MDs – n (control) = 267 RNs and 40 MDs – Convenience sample of RNs and MDs from all shifts, from eight sites; four sites received teamwork training and four did not	– Comparative non‐RCT descriptive study with a cross‐sectional correlational design using self‐reported validated quantitative questionnaires including the healthcare team vitality instrument a 10‐item, 5‐point Likert‐type scale survey, and revised nurse work index (NWI‐R), both previously validated in many health care settings – Descriptive analysis and t‐tests conducted using SAS	– Inclusion criteria for staff at both sites included that they had worked in ED for at least 6 months and were either full or part‐time – Staff demographics and ED experience align closely – 0.91 Cronbach's alpha > 0.84	– ED clinical staff who received teamwork training showed higher levels of staff perception of job environment, autonomy and control over practice. This included more positive perceptions by staff of access to resources and feeling like their opinions were more valued	– Interventional study exploring effects of a positive intervention on ED staff perception of WE	– ‡‡, ††No segregation of effects on MD versus RN – No data around response rate presented	– Training interventions can rapidly and positively affect staff perception of working environment, and this may also impact on patient care and safety, as well as staff turnover	50
9. Person, Spiva and Hart, 2013, USA	To examine the culture of an ED examining influences including stressful situations, pressure to perform and work‐life balance	– Included ED nurses, physicians, clinical care partners, technicians, customer servicers, leadership and support staff – n = 250 – n = 120 observation periods of 430 h – Interviews, n = 34	– §§Qualitative study using focused ethnographic exploration – Included a 16‐item demographic survey and informal and formal interviews, field notes, journaled identification of potential biases – Constant comparative method with external verification and calculation of cultural salience – Free listing responses	– Team member checks and meetings, reflexive journaling and audit trail including field notes, audiotapes, transcripts	– Culture primarily described by four categories; cognitive including teamwork and ability to multi‐task; environmental including limited physical space, poor work flow and overcrowding mixtures of acute and chronic stressors, technological limitations; linguistic including issues around barriers to communication and miscommunication and social attributes, siloing of knowledge and access, unprofessional behaviours, leadership (and staff) turnover, rites of passage	– Exploring a rich and wide range of staff perceptions – by direct observation and self‐reporting – Examining stressors within a working context – Data captured across a range of shift days and times – Enabling subtle influences on work place stress to be revealed	– Only a small portion of the ED culture revealed – Not easily replicated – No adjustment for experience or other demographic factors – Varied recording times – One ED site	– Management must value staff – Improve workflow processes and remove barriers – Development and training opportunities required for staff	100
10. Rasmussen, Pedersen, Pape et al., 2014, Denmark	To determine the relationships between staff perception of WE and the occurrence of adverse events	n = 124 – n = 98 ED nurses, n = 11 medical specialists and n = 15 junior doctors – One study site	– Quantitative descriptive using a cross‐sectional correlational design from a self‐reported validated quantitative questionnaire – Copenhagen psychosocial questionnaire job demands and influence components + demographics – Linear regression analyses + descriptive stats	– Validated scales providing quant parametric data – Response rate 91%	– Four of the five working scales included in the staff perception of working environment questionnaire returned ‘poor’ findings and were positively correlated with incidence of adverse events; poor team climate, poor inter‐departmental working relationships, poor safety climate, greater cognitive demands	– Data collected across clinical disciplines, demonstrates the clinical importance of staff perception of working environment for patient safety	– ‡‡, ††No description of survey follow up to encourage participation	– Ongoing assessment of adverse events in ED must be assessed in light of staff perception of the working environment	75
11. Lambrou, Papastavrou, Merkouris and Middleton, 2014, Cyprus	To examine nurses' and physicians' perceptions of professional environment and its association with patient safety in public EDs in Cyprus	n = 224 – n = 174 nurses and n = 50 physicians – All five possible public ED sites	– Quantitative descriptive study cross‐sectional correlational design from a self‐reported validated quantitative questionnaire including the Revised Professional Practice Environment (RPPE) Scale and (b) the Safety Climate Domain of the Emergency Medical Services Safety Attitudes Questionnaire (EMS‐SAQ) each with 4‐ to 5‐point Likert scale ratings – Demographic and professional experience data were also collected – Descriptive statistics, Mean differences were assessed using t‐tests and ANOVA, Bivariate association was evaluated using Pearson's correlation coefficients – Logistic and stepwise multiple regression modelling were undertaken to determine important relationship factor predictors	– 224/277 of possible participants (81% response rate) – 174/210 eligible nurses – response rate 83%, and 50/67 physicians – response rate 75% – The internal consistency of each of the RPPE sub‐scales was assessed using Cronbach's alpha coefficient	– Medical staff rated the professional practice environment slight more highly than nursing staff, particularly around ‘staff relationships’, ‘internal motivation’ and ‘cultural sensitivity’. While both groups rated teamwork highly, both groups also rated ‘control over practice’ as the lowest domain examined ‐Staff are highly motivated and in indicate that they value and practice team work	– Clear data collection period and well‐stipulated eligibility criteria – Reasonable gender spread; women (54%) of sample amongst both physicians and nurses (56% and 53%, respectively) – Direct comparison with staff from other hospital areas was possible – Previously undertaken research also enabled longitudinal comparison of data from nurses across three previous years	– ‡‡No description provided of other possible sites (private?) – RPPE Questionnaire was used in both medical and nursing staff, and some questions could have been perceived differently by the participants; impacting responses provided – A relatively low Cronbach's alpha was observed for two factors of the PPE as well as for EMS‐SAQ (safety domain)	– Improvements in professional environment can ultimately improve patient safety – Identified a need to focus on leadership interventions on the problematic areas to create an environment that is conducive to the delivery of safe and high‐quality care	100

^† Note: all survey completion was deidentified and voluntary, with appropriate accompanying ethical approval unless noted otherwise.

^‡ Data type (quantitative/qualitative) is identified in the study and/or on the basis of the analysis performed.

^§ Note: all survey and interview data are subject to potential prevarication bias and even response falsification. Additionally, the selections required in surveys are often ‘relative’ and so can be challenging ascertain consistently and reliably (‘soft’ responses). Additionally, there may be a response bias based on the psychological well‐being of participants (single point in time survey).

^¶ There were additional study findings not related to the focus of this review not reported here.

^†† Convenience (cross‐sectional) sampling and thus no causal inferences can be drawn.

^‡‡ No provision for open‐ended responses so participants' responses are constrained by study.

^§§ MMAT classification system

A&E, accident and emergency; hrs, hours; MD, medical doctor; MMAT, mixed methods appraisal tool; OD, Orthopaedics department; RN, registered nurse; RPPE, Revised Professional Practice Environment; SAS, statistical analysis systems; SD, standard deviation; SEM, standard error of the mean; SPSSx, statistics package for the social sciences; SSEM, Spanish Society of Emergency Medicine; SHOs, senior house officers; USA, United States; WE, work environment; WHOQOL, World Health Organization Quality of Life.