Abstract:-Aim. This paper is a report of the development and testing of the Self-Efficacy forPreventing Falls Nurse and Assistant scales.
Patient falls and fall-related injuries are traumatic ordeals for patients,family members and providers, and carry a toll for hospitals. Self-efficacy is animportant factor in determining actions persons take and levels of performance theyachieve. Performance of individual caregivers is linked to the overall performance ofhospitals. Scales to assess nurses and certified nursing assistants’ self-efficacy toprevent patients from falling would allow for targeting resources to increase SE,resulting in improved individual performance and ultimately decreased numbers ofpatient falls.
Four phases of instrument development were carried out to (1) generateindividual items from eight focus groups (four each nurse and assistant conducted inOctober 2007), (2) develop prototype scales, (3) determine content validity duringa second series of four nurse and assistant focus groups (January 2008) and(4) conduct item analysis, paired t-tests, Student’s t-tests and internal consistencyreliability to refine and confirm the scales. Data were collected during February–December, 2008.
The 11-item Self-Efficacy for Preventing Falls Nurse had an alpha of 0Æ89with all items in the range criterion of 0Æ3–0Æ7 for item total correlation. The 8-item Self-Efficacy for Preventing Falls Assistant had an alpha of 0Æ74 and all items had item total correlations in the 0Æ3–0Æ7 range.
The Self-Efficacy for Preventing Falls Nurse and Self-Efficacy forPreventing Falls Assistant scales demonstrated psychometric adequacy and arerecommended to measure bedside staff’s self-efficacy beliefs in preventing patientfalls.
Keywords: falls, hospital, instrument development, nursing, safety, self-efficacy
Patient falls are serious and long-standing problems in acutecare hospitals with published reports in the nursing literaturedating from 1979 (Walshe & Rosen 1979). Falls on hospitalmedical units have been reported to be as high as almost 3%(Vassallo et al. 2000). Falls are devastating to patients, familymembers, providers and the healthcare system with 34% offalls leading to injury and 6% to serious injury (Fischer et al.2005).There are abundant opinion reports on why hospitalizedpatients fall, tools to facilitate categorization of patients intohigh risk or low risk for falling, scales to identify specific riskfactors for falling, and strategies to prevent patients fromfalling. Yet, almost 30 years since the first paper on falls inhospitalized patients (Walshe & Rosen 1979), falls remain asignificant human ordeal for patients. Falls are also financiallycostly to hospitals as the Centers for Medicare andMedicaid Services do not reimburse hospitals for additionalcosts to care for fall-related injuries (Inouye et al. 2009).
Risks for falling:-
Being hospitalized increases the risk for falls (Evans et al.2001) secondary to the unfamiliar environment, acute illness,surgery, bed rest, medications, treatments and the placementof various tubes and catheters. Advancing age is another riskas patients aged between 76 and 85 years have a fivefold riskof falling compared with younger hospitalized patients(Halfon et al. 2001). Older adults are more likely to sustaininjuries with falls (Schwendimann et al. 2006, Krauss et al.2007) and injurious falls drive up hospital costs and length ofstay (Bates et al. 1995).
The higher risk older adults have for falling whilehospitalized is compounded by the fact that the elderly arethe fastest growing segment of the population. In general,hospitals have long been considered dangerous places forelderly patients (Cassel 2004). A single fall may result in afear of falling (Tinetti et al. 1990) that can begin a downwardspiral of reduced mobility, leading to loss of function andfurther risk for falls.
There are many published articles describing risk factorsfor falling and suggesting tools for assessing fall risk. In asystematic review of 47 such reports, two scales, the MorseFall Scale (MFS) (Morse 1997) and STRATIFY (Oliver et al.1997) fulfilled the criteria of prospective validation withadequate sensitivity/specificity values (Oliver et al. 2004).The MFS consists of six areas of risk and STRATIFY consistsof five factors independently related to a high risk of falling.When similar items are combined, there is a parsimonious listof eight fall risk factors: (1) recent history of falling or fall asa presenting complaint, (2) presence of secondary diagnosis,(3) need for ambulatory aid, (4) receiving intravenoustherapy, (5) gait problems or transfer or mobility problem,(6) impaired mental status or patient agitation assessed by anurse who knows the patient well, (7) need for frequenttoileting and (8) visual impairment. However, risk assessmentalone does not prevent falls. Interventions prevent falls.
Many resources have been devoted to fall prevention asevidenced in three decades of reports of fall preventionresearch. However, several systematic reviews published inthe last few years were largely inconclusive in their overallfindings of the efficacy of interventions to reduce fall rates inresidential care, nursing home and hospital facilities (Oliveret al. 2007, Coussement et al. 2008, Cusimano et al. 2008,Cameron et al. 2010). Post hoc subgroup analyses in thesereviews revealed that two individual components of multifactorialinterventions, vitamin D supplementation andsupervised exercise, were effective (Cameron et al. 2010).However, positive effects of these interventions are unlikelyto be realized during a typical 4-day hospitalization whenpatients are focused on recovering from surgery or from anacute medical problem. Results from the few studies on fallprevention programmes specific for hospitals have beenunsuccessful in identifying interventions to keep patientsfrom falling (Schwendimann et al. 2006, Krauss et al. 2008).
Why patients fall:-
Patients fall for a variety of reasons. Data from eight focusgroups [four RN and four certified nursing assistants (CNAs)]revealed six major reasons why patients fall: (1) inadequatepatient report, (2) lack of information access, (3) poorsignage, (4) unsafe environment, (5) lack of teamwork and(6) not involving the patient and family (Dykes et al. 2009).
Patients could have prevented some falls and staff could haveprevented other falls.
Patients:-When surveyed, older adults suggested that their balance,inattention and medical conditions were the most frequentreasons for falls (Zecevic et al. 2006). When interviewedabout a recent fall in the hospital, the major reason patientssaid they fell was the need to use the toilet coupled with loss ofbalance and unexpected weakness (Carroll et al. in press).Older females were not willing to face their risk of falling, andoften rejected fall prevention advice because they saw it as athreat to their autonomy and identity (McInnes & Askie2004).
Staff:-Bedside caregivers are the second line of defence for fallprevention, by helping patients do what they would do if theyhad the capability (Henderson 1977). Morse reported that78% of patient falls can be categorized as having an ‘anticipatedphysiological’ cause, meaning that a known physiologicalcause has placed the patient at increased risk forfalling (Morse 1997). Regrettably, after staff evaluate hospitalizedpatients’ risk of falling, very often, strategies thatwould counteract those risk factors are not carried out(Oliver et al. 2004). Interventions that are written on a careplan, but not implemented properly, do not prevent falls(Hendriks et al. 2008).
Background- A safety platform:We developed a model to depict how the patient, caregiversand environment are crucial to prevent falls. We modified theHome Safety/Injury Model (Hurley et al. 2004) used to guideresearch to make homes safer for persons with Alzheimer’sdisease (Horvath et al. 2005) and a working model to framepatient safety initiatives in hospitals (Kruger et al. 2006) toillustrate a safety platform for protecting patients from fallsand fall-related injuries (Figure 1). The patient is in thecentre. A patient care plan including current and accurate fallrisk status with associated tailored and feasible interventionsneeds to be easily and immediately accessible to the entirehealthcare team, patients and family (Dykes et al. 2009).The safety platform consists of two components, aprotective physical environment and bedside caregiverswho have both competence and SE beliefs in their abilityto prevent falls. A protective physical environment includeshaving immediate access to necessary assistive devices andthe presence of safe surroundings without unsafe conditions,e.g. call bell close at hand, a clear path to the bathroom,bed alarm if needed, durable call light pull-cords in thebathroom, intravenous poles with a wide base of support,and no frayed carpet, hazardous floor treads, or floorwashing when patients are likely to be in that area (Krugeret al. 2006).
Bedside caregiver competence means that RNs and CNAshave the knowledge and skills needed to carry out the careplan. RNs and CNAs must know (1) the patient’s specific riskfactors for falling, (2) explicit interventions to decrease,manage or ameliorate those risks and (3) how to carry outthose prescribed interventions. RNs and CNAs must beproficient in the actions they take to prevent patient falls.However, knowing exactly what to do and having thecapacity to do it capably are insufficient in preventing falls.RNs and CNAs must expend the degree of effort required andbe persistent in carrying out necessary actions to preventpatients from falling. Self-efficacy beliefs determine themagnitude of effort persons will employ and how long theywill persevere in their actions, in this case, to prevent patientsfrom falling.
Self-efficacy and patient falls:-
Self-efficacy:-The self-efficacy (SE) concept was first proposed by Bandura(1977) to understand why persons would engage in certainbehaviours to achieve their goals. SE is the major concept insocial cognitive theory (Bandura 1986) and is concerned withthe actions that people take rather than the outcomes that areexpected to follow. The first scales to operationalize andmeasure SE were used to study phobic behaviour(Banduraet al. 1977, 1982), hypertension (Rudd et al. 2004), chronicillness management (Lorig et al. 2001) and caregiving(Steffen et al. 2002). Bandura (1986) suggested that SE was apowerful psychosocial variable capable of predicting the enactment of health-related behaviours, which several decades of research have so confirmed. SE research, summarized in nine meta-analyses from diverse methodological and analyticstrategies, has affirmed the central role of SE in influencingactions (Bandura & Locke 2003).
SE is a belief in one’s capability to carry out actionsnecessary to meet given situational demands. RNs andCNAs, who have high levels of SE for preventing falls, willwork hard and persist in carrying out necessary actions toprevent patients from falling. Persons who perceive themselvesas efficacious are confident in their ability and expendsufficient effort to achieve a certain level of performance.This performance, when well executed, results in successfuloutcomes. Findings from a meta-analysis of 114 studies ofSE for work-related performance revealed that SE waspositively and strongly related to work performance, moderatedby task complexity – so that the relationship is thestrongest for lower levels of task complexity (Stajkovic &Luthans 1998). Actions required to prevent patients fromfalling are not complex; e.g. the patient needs two helpers totransfer from bed, requires a bed-alarm, and voids by usinga commode at the bedside. SE for the work-relatedperformance of carrying out actions to prevent falls isimportant because bedside RNs and CNAs are in keypositions to prevent falls and thus, fall-related injury (seeFigure 2).
Determinate of self-efficacy:-
Persons develop their SE beliefs from four hierarchically orderedsources (Bandura 1986). (1) Enactive attainment is amastery experience, ‘success breeds success’. (2) Vicariousexperience is a process of self-comparison to a peer. (3)Verbal persuasion is listening to something that alters one’sconfidence. (4) Physiological arousal is a physical indicatorgiven by the person’s own body. Fall prevention examples areused in Figure 2 to illustrate the four determinants of SE.RN and CNA SE for fall prevention was the onlycomponent of the safety platform (Figure 1) lacking in aquantification method. For this study, SE for preventing fallsmeans confidence in one’s ability to prevent patients fromfalling.
The study- Aims:The purpose of this study was to develop and test the SelfEfficacy for Preventing Falls – Nurse (SEPFN) and SelfEfficacy for Preventing Falls – Assistant (SEPFA) scales toallow hospital leaders to quantify bedside staffs’ SE beliefs inpreventing patient falls.
Methodology:-Four phases of investigation were carried out to develop andtest individual items and the SEPFN and SEPFA scales.Procedures based on classical measurement theory (Waltzet al. 2005) were employed to assure empirical, conceptualand psychometric adequacy. Established criteria were followedto ensure that both individual items and the SEPFNand SEPFA scales would be: (1) empirically grounded (Tinettiet al. 1990), (2) judged to have content validity (CV; Waltzet al. 2005), (3) accepted by RNs and CNAs (Wyatt & Altman 1995) and (4) reliable by meeting accepted standardsof internal consistency (Nunnally & Bernstein 1994).
Ethical considerations:-Institutional Review Board (IRB) approval was received atthe original four sites to conduct group interviews and to useweb-based and paper surveys. Separate IRB approval wasobtained at the additional sites for the survey phase usingelectronic dissemination via institutional listservs.Each participant in the focus groups obtained verbalconsent. Implied consent was provided for completing thesurveys. Survey instructions included the sentences: ‘Thisproject has received approval from the Human ResearchCommittee. Responding to these items indicates that yourecognize this is a research project in which you havevolunteered to participate. You may skip any items youchoose not to answer’.
Four study phases:-Participants from four acute hospitals, two urban academicmedical centres (800–1000 licensed beds) and two suburbanteaching hospitals (250–500 licensed beds) in the samehospital system, took part in phases one, three and four.These non-profit hospitals had no religious affiliation. Threeadditional hospital sites (non-profit, no religious affiliation,250–500 licensed beds) were included in phase four (reliabilitytesting). There were missing work site data on 199 RNparticipants and 171 CNA participants.
Phases one and three were qualitative descriptive componentsof the overall study. Participants were recruited frommedical units in the four acute hospitals with mean fall rateshigher than the hospital mean. RN and CNA potentialparticipants were identified by nursing leadership whorecommended potentially ‘good informants’, RNs and CNAsnurses who had both experienced the phenomena (in thiscase, being successful or not in preventing patients fromfalling) and could articulate their experiences (Morse 1987).Additional participants were recruited by flyers, e-mail andpersonal invitation. Informed consent was provided.
Phase one-Data (collected in October, 2007) obtained from four RNand four CNA audio-taped focus style group interviews(Dykes et al. 2009) were used to construct the qualitativedatabase. A combination of pre-planned questions, e.g. ‘Howdo you know if a patient is at risk for falling?’ ‘How do youknow what to do to prevent patient falls?,’ requests forclarification, e.g. ‘help me understand’ and ‘tell me moreabout’, and group discussion was used.
Interviews were transcribed verbatim into a word processingpackage, reviewed/corrected for transcription accuracyand removal/masking of any identifying characteristics ofpatients or providers, and converted into NVIVO for codingand support of qualitative analysis. We used a two-personconsensus for identification of participants’ text to beconsidered for the item pools following basic content analysismethods (Miles & Huberman 1994).
Phase two-Three nurse members of the research team (PD, DC, AH)with knowledge of professional practice issues surroundinghospital-based care and an understanding of the complex,technical and interdisciplinary care required by patients inwhich preventing falls is one of many safety concerns, identifiedstatements to be considered for the SEPFN and SEPFA(December, 2007). Guided by the definition of self-efficacyfor preventing falls, items were combined, refined, and editeduntil consensus was reached for 37 SEPFN and 33 SEPFAitems.
Phase three-A second series of four RN and four CNA focus groups wereheld (data collected in January, 2008). Participants used afour-point scale to rate each item for CV and relevance withthe definition of SEPF. As the paired t-test statistic was to beused in phase-four, an additional assessment was conducted totest participants’ understanding of and ability to followinstructions to create a private identifier, called the ‘SUM’ tomatch participants’ surveys. The ‘SUM’, used in our previousresearch with only nurses (Hurley et al. 2006, 2007), is thelast four digits of participants’ social security number added tothe last four digits of participants’ home telephone number.Participants were given verbal instructions and score sheetsfor rating items’ CV and writing their ‘SUM’. In the CNAfocus groups, there were several participants who could notcomplete the ‘SUM’ and additional participants who did nothave the English reading capacity to complete the CV ratingform independently. Refinement of items and instructionswas guided by group discussions to assure credibility (Wyatt& Altman 1995). Learning empirically from potential endpointusers, items and scale administration procedures wererevised and found acceptable.
The 90% average congruency percentage standard wasused to retain items. A 21-item SEPFN and 22-item SEPFAresulted and constituted the prototype scales. The confidentialindicator was expanded to allow respondents to select the‘SUM’ or a familiar word they would remember; e.g. the street where they grew up, favourite pet, etc. We learned that CNAs with low English literacy would need an alternative to SEPFA self-administration on a website to take part in phase-four.
Phase four-Reliability testing was accomplished in this final phase (datacollected April–August, 2008). A six-point Likert scoringsystem anchored with descriptors (completely confident)and (not at all confident) was used. Two identical SEPFNand SEPFA prototypes, with all positively worded items (tobe reverse-scored) and instructions for creating a confidentialidentifier using a word or numbers, were uploaded toSurveyMonkey, a web-based survey tool. The time-1 linkwas disseminated to collect data through electronic disseminationvia institutional listservs. Print copies wereprovided as requested and research staff uploaded the datainto the SurveyMonkey database. A week later, the time-2link was e-mailed and print copies were distributed asrequested and uploaded to collect data for examining re-teststability.
RNs and CNAs who provide direct bedside care weresought for reliability testing using a snowball samplingmethod. A recruitment email with the link to the T-1 SEPFNand SEPFA surveys and instructions was sent to collaboratorsat the four original sites, to other local hospitals and tocontacts through the Health Information Management SystemSociety Nursing Informatics Task Force. The email wasforwarded with the survey link using local listservs andprinted copies of the scales were provided to allow CNAs theoption of taking the SEPFA on paper privately or receivingassistance to use the paper version.
Data from printed versions were entered onto the form intothe SurveyMonkey database. The complete data sets weredownloaded from SurveyMonkey and converted to SPSS(Chicago, IL, USA) for the analysis.
Decision rules:-Decision rules were established to guide the analysis to retainonly well performing individual items to assure both reliableitems and parsimonious and psychometrically adequate scalesthat could detect overall differences in SE for preventing falls.T1 groups (G) were randomly divided into two (G1, G2) toexamine the data and make item reduction decisions with G1and evaluate psychometric properties with G2. A data set ofparticipants with identical confidential identifiers in T1 andT2 (matched) was used to examine item stability. Threeiterative decision rules were followed to delete poorlyperforming items if: (1) ‡80% participants (G1 data sets)did not score the item (not credible) or scored ‘strongly agree’(lack of differentiation capacity), (2) <0Æ5 Pearson correlation (matched data sets) and/or the probability of the t-value <0Æ1 (not highly correlated and/or dissimilar means – indicating lack of item stability evidenced by different scores for the same item when responded to at a 2-week interval) and (3) internal consistency reliability examination by itemtotal correlation (G1 data sets) of <0Æ3 (poor fit) or >0Æ7(redundant).
After deleting poorly performing items following thesequence of the decision rules described above, both theretained items and SEPFN and SEPFA were evaluated. Meanscores were used; the theoretical range is 1 (not at allconfident) to 6 (completely confident) in ability to preventpatients from falling. Scales were evaluated for internalconsistency (G1) and stability (matched data sets). The finalexamination consisted of a Student’s t-test to compare G1and G2 data sets. SEPF scales’ descriptive data and variabilitywere computed using the complete T1 data sets.
Sample/participants:-A total of 36 RNs participated in phases one and three and562 RNs responded to the T-1 survey, of whom 73 providedconfidential identifiers and were included in the matched dataset. A total of 27 CNAs participated in phases one and threeand 269 CNAs responded to the T-1 survey, of whom 83provided confidential identifiers and were included in thematched data set. Demographic characteristics are providedfor the total T1 RN and CNA respondents only as matchedparticipants were drawn from those groups and there were nodifferences between those participants who provided aconfidential identifier and those who did not. Table 1compares RN study participants with the most recent surveyof United States of America (USA) RNs (U.S. Department ofHealth and Human Services, HRSA, BuHP, & Division ofNursing, 2010) and Table 2 compares CNA study participantswith available CNA national data (Squillace et al.2009).
A greater percentage of RN study participants in all phaseshad a baccalaureate degree and were younger than RNsemployed in US hospitals. A greater percentage of CNA studyparticipants in phases one and three were not Caucasian,were older and had more experience at their work site thanCNA participants in phase four had.
Instruments/results:-Ten SEPFN items were deleted because they did not achievethe criteria of decision rules one and two by not beingcredible, lacking differentiation capacity, or being unstable.
Discussion:-Many aspects of preventing falls in acute care hospitalspresent some degree of difficulty. Patients are not admitted tohospitals to be prevented from falling, but are admitted andtreated for health problems serious enough to require acutecare hospitalization. Preventing patients from falling is justone of many actions staff perform as they provide care andvery comprehensive therapeutic interventions to extremely illpatients who typically have several comorbid conditions.
Limitations:-Phases one and three were conducted with RNs and CNAsfrom one healthcare system. However, the four units wherestudy RNs and CNAs worked provided care for sick adultmedical patients who were presumably similar to medicalpatients cared for in USA acute care hospitals vs. patientsrequiring speciality/intensive care.
Non-probability sampling methods were employed torecruit participants in all phases. The sample may not berepresentative of all bedside RNs and CNAs. Limitationsassociated with internet surveys are well known, includingsampling bias and inability to calculate a response rate(Wyatt 2000).
There were age, gender, race/ethnicity and educationaldifferences between RNs in the two qualitative phases, thereliability testing survey phase, and employed in USAhospitals in 2008 (U.S. Department of Health and HumanServices et al., 2010). We could find no data on therelationship of gender, or race/ethnicity or educational leveland self-efficacy for any type of work-related performance.Although SE has been found to influence academic achievementin school (Bandura et al. 1996), it remains unknown if ahigher level of education would influence work-related SE.It is known that patient outcomes are better for thosehospitals with high rates of nurses with a baccalaureatedegree (Aiken et al. 2003, Kutney-Lee & Aiken 2008). Werecognize this educational difference, but believe that studyresults are generalizable to hospital-based RNs because wewere not seeking the application of a complex nursingjudgment, but SE beliefs in preventing patients from fallingand SE for work-related performance are stronger whenrelated actions are not complex (Stajkovic & Luthans 1998).CNAs in the two qualitative phases were older, more likelyto be male and not Caucasian, and had more work experienceat their current hospital than CNAs in the reliability testingsurvey phase and CNAs employed in USA nursing homes in2004 had (Squillace et al. 2009). However, we are not awareif these demographic characteristics influence SE for fallprevention.
Conclusion:-Suggestions for using the SEFPN and SEPFA scalesBoth individual items and the scale totals can be used toassess RNs and CNAs SE for preventing hospitalizedpatients from falling and should be tested with RNs andCNAs in other settings where patients may fall. Forexample, the item ‘I am confident in my ability to preventpatients from falling because we all work together as ateam’ is in both SEPFN and SEPFA scales, has satisfactorypsychometric properties, and could be one way to evaluatethe efficacy of a fall prevention programme that incorporatedteam building. If health information technology is used to streamline fall risk communication for nurses, the items concerning confidence in ability because of having easy access to information about why a patient is at risk to fall and how to prevent a patient from falling could be used as an evaluation mechanism.
Additional suggested research:-
Testing the SEPFN and SEPFA scales-We recommend that additional psychometric testing beperformed with larger samples in different hospitals and insettings such as long-term/residential care settings whereolder patients reside, to ensure that both individual items andthe scales are relatively stable. The elderly are at more riskfor falling while hospitalized and suffering greater negativeconsequences from falling than younger patients. Olderpatients are living longer lives and have very complex healthneeds, and there may not be adequate numbers of healthcareproviders with the knowledge and skills to care for themadequately (Institute of Medicine, 2010). The recent Instituteof Medicine report, ‘Retooling for an aging America: Buildingthe health care workforce’ (Institute of Medicine, 2010),stated that all providers need to have the core competenciesin caring for older persons during general training, lifelong,and when needed. The SEPFN and SEPFA are two new resourcesto use to promote the goal of fall and injury preventionand should be examined in all sites where RNs andCNAs provide care.
Additional research:-We recommend that the safety platform model developed forthis project (Figure 1) be examined. The two new scales canmeasure RN and CNA SE for fall prevention. RN and CNAcompetency can be tested by standardized evaluations ofknowledge, e.g. RNs at the health system where phases oneand three were conducted must demonstrate accurate use ofthe MFS (Morse 1997) annually. The degree to which theenvironment is protective can be examined by using patientsafety walk rounds (Frankel et al. 2003). Medical recordaudits can identify the completion of fall risk assessments andsubsequent fall prevention interventions and the degree towhich interventions are carried out. Particular attentionshould be given to the hospitalized elderly as they are at greatrisk for falling and are the fastest growing segment of thepopulation.
Interventions examined in previous fall prevention researchthat looked promising, but did not achieve statistical significancein the hospital setting for reducing falls, should bere-tried with the addition of RN SE for preventing falls. Forexample, the results of a recent evaluation of a multifactorialfall prevention programme were attributed to considerablediscrepancy between the intervention planned by theresearchers and what actually was carried out by thosepersons implementing the experiment (Hendriks et al. 2008).It is important that the persons who implement fall preventioninterventions maintain the necessary degree of effortrequired to carry out the protocol accurately and persist intheir actions to do so.
In summary, the SEPFN and SEPFA scales are suggestedfor use in any site where fall prevention is a goal. We hopethat suggestions for future research will help solve the seriousproblem of falls in hospitalized patients.
Author contributions:-PCD, DLC and ACH were responsible for the study conceptionand design, drafting of the manuscript. PCD, DLC, KM,LC and LZ performed the data collection. PCD, DLC, ACH,SRL and BM performed the data analysis. KM, SRL, LC, LZand BM made critical revisions to the paper for importantintellectual content. LZ provided statistical expertise. PCD,DLC, ACH and BM obtained funding. KM and LZ providedadministrative, technical or material support. PCD and ACHsupervised the study.
Conflict of interest-No conflict of interest has been declared by the authors.
References-Aiken L.H., Clarke S.P., Cheung R.B., Sloane D.M. & Silber J.H.(2003) Educational levels of hospital nurses and surgical patientmortality. Journal of the American Medical Association 290,1617–1623.Bandura A. (1977) Self-efficacy: toward a unifying theory of behavioralchange. Psychological Review 84, 191–215.Bandura A. (1986) Social Foundations of Thought and Action: ASocial Cognitive Theory. Prentice Hall, Englewood Cliffs, NJ.Bandura A. & Locke E.A. (2003) Negative self-efficacy and goaleffects revisited. Journal of Applied Psychology 88, 87–99.Bandura A., Adams N. & Beyer J. (1977) Cognitive processesmediating behavioral change. Journal of Personality and SocialPsychology 35, 125–139.Bandura A., Adams N. & Beyer J. (1982) Microanalysis of actionand fear arousal as a function of differential levels of perceivedself-efficacy. Journal of Personality and Social Psychology 43, 5–21.
Let’s block ads! (Why?)