Leca

MaEma Insb Cenc L-Bd De

International Journal of Nursing Studies xxx (2012) xxx–xxx

A R

Artic

Rece

Rece

Acce

Keyw

Imp

Qua

Nur

Safe

*

G Model

NS-2099; No. of Pages 10

Plac

002

http

vels and correlates of implicit rationing of nursing care in Swiss acutere hospitals—A cross sectional study

ria Schubert a,b, Dietmar Ausserhofer a, Mario Desmedt a, Rene Schwendimann a,manuel Lesaffre c,d, Baoyue Li d, Sabina De Geest a,*

titute of Nursing Science, University of Basel, Switzerland

tre of Clinical Nursing Science, University Hospital Zurich, Switzerland

iostat, Katholieke Universiteit Leuven, Belgium

partment of Biostatistics, Erasmus MC, Rotterdam, The Netherlands

T I C L E I N F O

le history:

ived 3 January 2012

ived in revised form 11 September 2012

pted 13 September 2012

ords:

licit rationing of nursing care

lity nurse practice environment

sing resources

ty climate

A B S T R A C T

Objectives: (1) To describe the levels of implicit rationing of nursing care in Swiss acute

care hospitals; (2) to explore the associations between nine selected potential rationing

predictors and implicit rationing of nursing care.

Design: Cross sectional multi-center study.

Settings: A quota sample of 35 acute care hospitals from the German, French and Italian

speaking regions of Switzerland participating in RN4CAST (Registered Nurse Forecasting) study.

Participants: 1633 registered nurses working in randomly selected medical, surgical or

medical–surgical units.

Methods: Implicit rationing of nursing care, i.e., the withholding of any of 32 necessary

nursing measures due to a lack of nursing resources, was measured using the revised Basel

Extent of Rationing of Nursing Care (BERNCA) instrument. Nine potential rationing

predictors, e.g., staffing and resource adequacy, patient-to-nurse ratio, nurse education,

and confounding variables, e.g., nurse gender and age, hospital typology, were assessed

with validated instruments or single items of the RN4CAST study. Descriptive statistical

procedures were used as appropriate. Three level regression models were used to

investigate the effect of the selected nine predictors on rationing at the nurse, unit and

hospital levels.

Results: Ninety-eight percent of the participating nurses reported that, in their last seven

working days, they had to ration at least one of the 32 nursing tasks listed in the BERNCA.

The mean rationing level of 1.69 (SD = 0.571) indicates that on average the nurses reported

‘rarely’ being unable to perform the nursing tasks listed in the BERNCA. Multilevel

regression analysis confirmed two of the nine tested predictors: better unit level staff

resource adequacy and a more favorable hospital level safety climate were both

consistently significantly associated with lower rationing levels. Counter to our

assumptions, the other two nurse practice environment dimensions, the three workload

measures, nurse experience and nurse education were not associated with rationing.

Conclusions: Rationing frequency varied among the 32 BERNCA items, indicating differing

prioritizations of necessary nursing tasks. The identified rationing predictors, staff

resource adequacy and safety climate, can determine starting points for interventions, i.e.,

proactive changes to improve staff resource adequacy when rationing exceeds predefined

thresholds, increasing the risk of negative effects on patient outcomes.

� 2012 Elsevier Ltd. All rights reserved.

Corresponding author.

E-mail address: sabina.degeest@unibas.ch (S. De Geest).

Contents lists available at SciVerse ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

ease cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

0-7489/$ – see front matter � 2012 Elsevier Ltd. All rights reserved.

://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx2

G Model

NS-2099; No. of Pages 10

What is already known about the topic?

� In acute hospital settings even low levels of implicitrationing of nursing care have been proved to be a strong,independent predictor of patient outcomes, such asmedication errors, falls, pressure ulcers, critical incidentsand nosocomial infections.� Implicit rationing of nursing care appears to partially

explain the effects of the practice environment andpatient-to-nurse ratio on outcomes. However, its con-tributing factors are not well understood.

What this paper adds

� Variability in rationing levels among the measurednecessary nursing tasks indicates that, when it is notpossible to provide all tasks to all patients, nursesprioritize their care functions. Specifically, tasks relatedto medical, technical, and therapeutic treatments wereless frequently rationed than those in the areas of caringand support.� The multilevel analyses results contribute to a clearer

understanding of relevant correlates of implicit rationingof nursing care in acute care hospitals, and may thereforecontribute to the development of improved interventionapproaches.

1. Introduction

As with other healthcare professionals, nurses’ func-tions in hospitals are compromised when cost contain-ment/regulation strategies lead to resource shortages.Limited resources (i.e., staffing and skill mix) force nursesto prioritize necessary nursing care and might result inpatients not receiving needed nursing measures (Aikenet al., 2001; Kuenzi and Schaer-Moser, 2002; West et al.,2007). Implicit rationing of nursing care – the provision ofnursing care levels below what the responsible nursesjudge necessary to reach the desired patient outcomes orprotect patients from negative events or conditions, suchas falls or pressures sores – is a prominent factor of overallhealthcare effectiveness and costs.

To our knowledge, implicit rationing of nursing careand its association with patient outcomes was system-atically studied for the first time in the Rationing ofNursing Care in Switzerland (RICH Nursing) study (Schu-bert et al., 2007, 2008, 2009), the Swiss extension of theInternational Hospital Outcomes Study (IHOS) (Aiken et al.,2002a,b). The results of this study show that rationing hassignificant associations with inpatient complications (e.g.,falls, pressure ulcers, nosocomial infection, critical inci-dents, medication errors) and mortality rates, underliningits relevance as a system factor (Schubert et al., 2008, 2009,2012). Because of the shown relevance of this factor and inorder to know where to intervene when rationing exceedsthe threshold levels defined in the RICH Nursing study, is itimportant to develop a clearer understanding of itsmechanism and potential rationing correlates.

In order to investigate the potential predictors ofrationing this study builds on the conceptual framework ofimplicit rationing of nursing care developed for the RICHNursing study, including the definition. To position

rationing in the broader context of system factorsinfluencing patient and nurse outcomes, this frameworkwas adapted from that used for the IHOS. To measureimplicit rationing of nursing care, the Basel Extent ofRationing of Nursing Care (BERNCA) instrument wasdeveloped (Schubert et al., 2007, 2008, 2009).

1.1. Conceptual framework of implicit rationing of nursing

care

Based on the available evidence and the definition ofimplicit rationing in medicine, implicit rationing of nursingcare was defined as ‘‘the withholding of or failure to carryout necessary nursing measures for patients due to a lackof nursing resources (staffing, skill mix, time). Consideringthat rationing of nursing care also occurs at the meso- andmacro levels of healthcare, however, this definition isrestricted to rationing at the micro level, i.e., resourceconstraints at the level of the individual nurse. Within thisdefinition, ‘‘necessary nursing measures’’ refers to a groupof nursing tasks and actions which are accepted – both byclinical consensus and by the attending nurse – asimportant for a patient to achieve the desired outcomes.Such tasks and actions can be classified according to thefollowing aims: surveillance, therapy, support, preventionor prophylaxis, activation or rehabilitation, educationaland instructional measures, and measures related to theapplication, documentation, and adaptation of nursingprocesses. The individual priorities of these measures arebased on the nurse’s assessment of the patient’s healthcareneeds (number of nursing problems identified, desiredoutcomes, and evaluated effectiveness of the care plannedduring the nursing process), scientific evidence, reflectedexpert knowledge and patient preference (Schubert and DeGeest, 2003). Further, the necessity of each measure isprescribed by nursing standards and local and nationalguidelines and procedures (see Fig. 1). The frameworkshows the assumed relationships between organizationalsystem factors, individual nurse and patient character-istics, rationing and patient and nurse outcomes. Theoutcomes are not the focus of this study (see Fig. 1).

As shown in the framework, implicit rationing ofnursing care occurs at the patient-to-nurse interface andis a product of clinical decision-making and criticaljudgment. In general, decision-making is seen as a complexprocess involving the following steps: information collec-tion, problem identification, consideration of alternativestrategies, and selection of optimal actions. Factorsinfluencing decision-making include educational level,compliance with unit-based practice, hospital culture,leadership and responsibility, all of which are influencedby experience, confidence, willingness, the nurse’s abilityto take responsibility, the patient’s condition (stable,unstable), and the available resources (real or impendinglack of resources) (Hancock and Easen, 2006). Factorsinfluencing priority-setting in clinical nursing practiceinclude characteristics of the nursing care work environ-ment (e.g., organization of nursing practices at the wardlevel, nurses’ autonomy and responsibility), amount ofavailable time and resources (e.g., number of staff, skillmix, equipment), philosophy of care and aims of the

Please cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissacute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

orgnatpattionmeexp(Bu201impreleimbres

assassrelaandKalfactticabeethenurresto-nAlsInslinkresUniresfou

Fig.

et a

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx 3

G Model

NS-2099; No. of Pages 10

Plac

anization (e.g., nursing care standards, local andional policies, regulations, guidelines, safety culture),ients’ characteristics and the acuteness of their condi-s (e.g., number and urgency of health problems),

dical treatment plans, nurse characteristics (e.g.,erience and educational level of the nursing team)cknall, 2000; Hendry & Walker, 2004; Kalisch et al.,1). It can be assumed that all of these factors influencelicit rationing of nursing care, but that the mostvant are nurse staffing resource adequacy, the balance/alance between patient care needs and nursing

ources (workload), respectively.Rationing of nursing care or related concepts have beenessed in a number of studies focusing primarily onessing the level of rationing. To a limited extent thetionships between implicit rationing of nursing care

other factors were investigated (Aiken et al., 2008;isch et al., 2011; Rochefort and Clarke, 2010). So far, theors influencing rationing, as described in the theore-l framework, and its underlying mechanism have notn studied systematically. Still, preliminary results of

RICH Nursing study suggest associations with these work environment, particularly the staffing andource adequacy (r = �0.46) dimension and the patient-

urse staffing ratio (r = 0.14) (Schubert et al., 2007).o, using the Neonatal Extent of Work Rationingtrument (NEWRI), Rochefort et al. presented evidenceing the nurse work environment dimension of ‘nurse

ource adequacy’ to rationing in Neonate Intensive Carets (NICUs). In their study, more favorable staffing and

ources were associated with 5.7–11% decreases in allr measured rationing dimensions: life support and

technically oriented care; parental teaching and supportand infant comfort care; patient surveillance; and carecoordination and discharge planning. They also found anassociation between the nurse work environment ‘nursephysician relationship’ dimension and the rationingdimensions of life support and technically oriented careand patient surveillance (Rochefort and Clarke, 2010).

However, as both studies used small, non-representa-tive samples, exploring and confirming their low butsignificant correlation between implicit rationing andorganizational factors such as the work environment,the patient-to-nurse ratio, and the as yet unstudied effectof patient safety climate on rationing requires furtherresearch. Therefore, the purposes of this study were (1) todescribe the levels of implicit rationing of nursing care in aquota sample of Swiss acute care hospitals and (2), toexplore the assumed associations between the quality ofthe nurse work environment dimensions, patient-to-nursestaffing ratio, number of patients requiring support in allADLs, number of patients requiring hourly or morefrequent monitoring, patient safety climate, and nurseexperience and education and implicit rationing of nursingcare. These factors were selected based on the conceptualframework developed for the RICH Nursing study and thedata collected for it in the context of the RN4CAST study.

The following hypotheses guided this study: (H1) ahigher quality of the nurse practice environment, particu-larly the perceived staff resource adequacy, is related withlower levels of implicit rationing of nursing care; (H2) ahigher patient to nurse ratio, higher number of patientsrequiring support in all ADLs and/or required hourly or morefrequent monitoring is related to higher levels of implicit

Organizational va ria bles

•Budget

•Policy priorities

•Resou rce alloca tion

•Management structure

•Cul ture and climate

Patient va ria bles

•Type & severity of illness

•Co-morbidities & risk fac tors

•Number of health problems

•Medica l trea tment pl an

•Acut eness of patient condition

Nursing work environment

•Adequacy resource s / skill mix

•Interdisciplinary collabo ration

•Nu rsin g management

•Autonomy & respo nsibility

Patien t outcomes

Nurse outcomesPhilosophy of care

• Priority setting

• Cul ture values

• Nu rsin g ca re standards

• Local & national

guidelines, proce duresNurse va ria bles

• Experience (yea rs as a nurse)

• Educa tion

• Skills / knowledge

Nurses

(resource s)

Patient

(ca re needs)

Levels of implicit

rationing of

nursing care

Decision making

Clinical jud gment

Process of care

Workload

Ass essment Planning

ImplementingEvaluation

1. Theoretical framework with suggested relationships between selected factors and implicit rationing of nursing care (Aiken et al., 2002a; Schubert

l., 2007).

ease cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx4

G Model

NS-2099; No. of Pages 10

rationing of nursing care; (H3) a more favorable patientsafety climate is related to lower levels of implicit rationingof nursing care; (H4) higher nurse educational levels andincreased experience (years worked as a nurse) are relatedwith lower levels of implicit rationing of nursing care.

2. Methods

2.1. Design, setting and sample

This is a sub-study of the cross-sectional, multicenterRN4CAST (Registered Nurse Forecasting) study, whichinvolves hospitals and nurses from 12 European countries(Sermeus et al., 2011). The Swiss part, described here,includes a quota sample of 35 acute care hospitals from theGerman, French, and Italian language regions. Quotasampling was done based on geographical location (aminimum of one hospital per canton, two in large-areacantons), number of beds (at least 60 beds) and hospitaltype (general acute care hospitals providing medical andsurgical services). In order to achieve the desired numberof at least 50 nurses per hospital, three to four medical,surgical or mixed medical–surgical units per hospital wererandomly selected, whereas in smaller hospitals allmedical and surgical units were included. All registerednurses (n = 2280) working in direct patient care in theseunits were surveyed. Of these, 1633 nurses participated.

The original English questionnaires of the RN4CASTstudy, as well as the additional questionnaires used in theSwiss part of the study, were translated into German,French, and Italian following the modified Brislin transla-tion protocol for cross-cultural research (Jones et al., 2001;Sermeus et al., 2011). Additionally, a content validityrating, provided by a sample of 29 Swiss nursing expertsrepresenting Switzerland’s three language regions, wasperformed (Squires et al., 2012).

2.2. Study variables and measurements

2.2.1. Outcome variable

Implicit rationing of nursing care, defined here as thenumber of necessary nursing tasks for patients withheld orotherwise not performed in the last seven working days,due to inadequate time, staffing levels and/or skill mixes.This variable was measured with the revised version of theBasel Extent of Rationing of Nursing Care (BERNCA-R)instrument. This tool contains 32 Likert type questions(0 = not required, 1 = never, 2 = rarely, 3 = sometimes,4 = often). The construct validity (uni-dimensional struc-ture) and reliability (Cronbachs Alpha 0.94) of the BERNCAwere confirmed in the RICH Nursing and RN4CAST studies.Early results of the RICH Nursing study also indicated thatthe tool had good criterion validity (Schubert et al., 2007).

2.2.2. Predictor and control variables

Quality and elements of the nurse practice environment

were measured with the 32-item Practice EnvironmentScale of the Nurse Work Environment Index-Revised (PES –NWI-R). Based on magnet hospital research, this toolincludes five subscales: (1) nurse participation in hospitalaffairs, (2) staffing and resources adequacy, (3) nurse

foundations for quality of care, (4) nurse manager ability,leadership and support of nurses, and (5) collegial nursesand physicians relations (Aiken et al., 2002a,b; Sermeuset al., 2011). Using a 4-point Likert-type scale ranging from‘‘strongly agree’’ (=4) to ‘‘strongly disagree’’ (=1) nursesindicated the extent to which each element was present intheir current job. The reliability of the PES-NWI R, i.e.,Cronbach’s alphas for NWI-R, varied from 0.70 to 0.89(Lake, 2002; Schubert et al., 2007). Its predictive validityhas been confirmed in other recent studies (Aiken et al.,2008; Bruyneel et al., 2009)

Patient-to-nurse staffing ratio defined here as thenumber of patients on the unit in the last shift, dividedby the number of RNs on the unit in the last shift, wasmeasured with a single item from the RN4CAST ques-tionnaire. The number of patients on the respondent’s unitwho needed assistance with all activities of daily living

(ADLs) and the number requiring hourly or more frequent

monitoring or treatment were also measured with singleitems from the RN4CAST questionnaire. Nurses were askedto indicate the number of patients and nurses involved, aswell as the number of patients requiring assistance with allADLs or frequent monitoring in their last worked shift.

Reflecting a ‘‘visible feature’’ of an underlying mindfulsafety culture, patient safety climate was defined as thedegree to which nurses engaged in crucial patient safetybehaviors and practices on their units. Patient safetyclimate was measured with the nine-item Safety Organiz-ing Scale (SOS). Using a 7-point Likert type scale rangingfrom not at all (=1) to a very great extent (=7), nurses wereasked to indicate the extent to which they and theircolleagues engaged in each of the nine listed safetybehaviors. The validity and reliability of different languageversions of the SOS has been confirmed in different cross-cultural settings (Ausserhofer et al., 2012; Vogus andSutcliffe, 2007). For instance, based on confirmatory factoranalysis, the internal structure (e.g., comparative fitindices > 0.90, root mean square error ofapproximation < 0.08) and reliability (Cronbach’salpha > 0.79) support the SOS’s uni-dimensionality.

Nurse characteristics: Nurse experience – the number ofyears worked as a nurse in the current hospital, Educational

qualification – university degree (yes/no), Age – in years, Sex –female/male, Employment status – full-time (yes/no) or part-time (percentage). These data were gathered using singleitems from the RN4CAST questionnaire, for which respon-dents were asked to report the required information.

Hospital characteristics: Typology, i.e., large generalteaching hospitals (Level 1, university), large generalhospitals withsome teaching activities (Level 2), and general,mixed or community hospitals (Levels 3–5, respectively);and Size, i.e., number of beds, were either collected usingitems of the RN4CAST organizational variable questionnaire(Sermeus et al., 2011) or extracted from data published bythe Swiss Federal Office of Public Health (OPH) or the SwissFederal Statistical Office (FSO) (BAG, 2008).

2.3. Data collection

Approvals were obtained from 13 local ethic commit-tees responsible for the selected 35 hospitals. The data

Please cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissacute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

coll200anopermesen

2.4.

tivelevlyserelaraticonfollmunurabiandneeor mexpgenwe

levmoprebetandstruregparhosoutduapreinteassa thsigndevcor(SPUSAmu

3. R

3.1.

thalevme16335

hosisti

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx 5

G Model

NS-2099; No. of Pages 10

Plac

ection was consecutively conducted between October9 and June 2010. Nurses were surveyed voluntarily andnymously. In each involved hospital designated contactsons distributed the questionnaires to all nurses whot the inclusion criteria. Completed questionnaires weret back in a pre-paid envelope to the research team.

Data analysis

Variables were summarized using appropriate descrip- and explorative measures based on measurement

els and data distributions. Multilevel regression ana-s were used to test the hypotheses and to evaluate thetionship between selected potential correlates andoning. Based on the hypotheses formulated, theceptual model and the availability of data, theowing correlates were selected and included in theltilevel regression models as potential predictors: threese practice environment dimensions: nurse managerlity, staffing and resource adequacy, collegial nurses

physicians; patient-to-nurse staffing ratio; patientsding assistance with all ADLs; patients requiring hourlyore frequent monitoring; patient safety climate; nurse

erience and education. In the resulting models, nurseder, age and employment level, and hospital typology

re tested as potential confounders.Due to the natural clustering of the data (into threeels: hospitals, units, and nurses) multilevel regressiondels were used to investigate the effects of eachdictor on rationing. Intraclass correlation coefficientsween the three level structures of 3.6% (hospital level)

8.7% (nursing unit level) confirm the multilevelcture of the data and justify the use of multilevel

ression analysis. To build the models all predictors weretitioned, aggregated respectively at the nurse, unit andpital levels (Neuhaus and Kalbfleisch, 1998). Thecome variable rationing was aggregated at the indivi-l level. First, separate three level models for eachdictor group with unit and hospital level as randomrcept were developed, to test the hypothesized

ociations between each predictor and rationing. Second,ree level multiple random effects model with only theificant predictors based on the separate models wereeloped. The level of significance was set at a = 0.05. Norection for multiple testing was envisaged. SPSS 19SS for Windows, Rel. 19. 2010. IBM Inc., Armonk, NY,) and R package 1me4 were used to perform the

ltilevel analyses.

esults

Characteristics of hospitals and nurses

Of the 35 included hospitals, roughly half had fewern 200 beds. Similar numbers offered acute care serviceels 1 or 2. Of the 132 included hospital units, 48% weredical, 45% surgical and 7% mixed surgical/medical. The3 participating nurses (90% females) were on averageyears old and had worked five years in their currentpitals. Further details of nurse and hospital character-

3.2. Frequencies and levels of implicit rationing of nursing

care

Of the participating nurses, 98% reported that, in theirlast seven working days, they had to ration at least one ofthe 32 nursing task listed in the BERNCA. The meanrationing level of 1.69 (SD = 0.571) (on a scale with a 1–4theoretical range) indicates that, on average, nursesreported ‘rarely’ being unable to perform the nursingtasks listed in the BERNCA. Nurses reported that the mostfrequently rationed nursing tasks related to caring andsupport or documentation (e.g., ‘offer of emotional orpsychosocial support to patient’, ‘assessment of newlyadmitted patients’). Nursing tasks such as continencetraining, patient education and training, preparation ofpatients for tests and therapies were less frequentlyrationed (Table 2).

3.3. Summaries of the rationing predictor variables

Summaries of the quality of the nurse work environ-ment (three PES-NWI-R dimensions), the patient-to-nurseratio, the number of patients requiring support in all ADLs,patients requiring hourly or more frequent monitoring,and safety climate, aggregated at the unit level, arepresented in Table 3. The nurses neither agreed nordisagreed that staff resources on their unit were adequate(mean = 2.52), but assessed their nurse managers’ abilitiesand collaboration between nurses and physicians in theirunits more favorably. On average, respondents reportedhaving been responsible for eight patients on their lastshift. Of these, an average of four patients needed supportwith all ADLs and two requiring hourly or more frequently

Table 1

Summary of hospital and nurse characteristics.

Variable

Hospital characteristics (N = 35)

Size

<200 beds – % (N) 51 (18)

200–600 beds – % (N) 32 (11)

>600 beds – % (N) 17 (6)

Typology

Large general hospital (level

1, university) – % (N)

11 (4)

Large general hospital (level 2) – % (N) 40 (14)

General hospital mixed specialist

(level 3) – % (N)

23 (8)

General hospital community

(level 4) – % (N)

20 (7)

General hospital community

(level 5) – % (N)

6 (2)

Unit characteristics (N = 132)

Medical units – % (N) 48 (64)

Surgical units – % (N) 45 (59)

Mixed surgical and medical units – % (N) 7 (9)

Nurse characteristics (N = 1633)

Age-mean (SD) 35 (9.89)

Sex: female – % (N) 90 (1469)

Employment: full time – % (N) 49 (768)

Education: university degree (BSN/MSN) – % (N) 10 (151)

Years working – median (IQR)

As a nurse 8.00 (14.81)

In this hospital 5.00 (10.17)

standard deviation; IQR = interquartile range.

cs can be found in Table 1. SD =

ease cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx6

G Model

NS-2099; No. of Pages 10

monitoring. Finally, the patient safety climate scoresindicated that the participating RNs were involved to aconsiderable extent in patient safety behaviors andpractices on their units (mean: 4.98, SD 0.544).

3.4. Hypothesis testing

As hypothesized (H1), a significant association was foundbetween the nurse practice environment ‘‘staff resourcesadequacy’’ dimension and implicit rationing of nursing care.More favorably estimated unit level staff resource adequacy

was associated with lower levels of implicit rationing in bothseparate model (p-value 0.001) (Table 4) and adjustedmodels (p-value 0.042) (Table 5). The other two nursepractice environment dimensions, ‘nurse managementability’ and ‘collegial nurse physicians’, were not, asproposed, significantly associated with rationing, either inthe separate (p-value 0.062–0.866) (Table 4) or in theadjusted model (p-value 0.186–0.905). Of the other threeorganizational system factors (patient tonurse ratio, numberof patients requiring support in all ADLs, and patientsrequiring hourly or more frequent monitoring) (H2) only the

Table 2

Levels and frequency of implicit rationing listed by BERNCA question.

No. Questions/items BERNCA revised Median

(IQR)

Often

(%)

Sometimes

(%)

Rarely

(%)

Never

(%)

Not

required (%)

1. Sponge bath 1.00 (1.0) 1.7 13.6 25.8 48.8 10.1

2. Partial sponge bath 1.00 (1.0) 1 11.8 27.2 56.2 3.7

3. Skin care 1.00 (1.0) 2.1 14.8 33.4 48.0 1.7

4. Oral hygiene 2.00 (1.0) 5 18.5 29.8 36.4 10.3

5. Dental hygiene 2.00 (1.0) 4.5 17.6 32.6 37.9 7.4

6. Assist food intake 1.00 (1.0) 4.2 10.6 22 46.4 16.8

7. Mobilization 2.00 (2.0) 5.9 28 36.7 27.4 2

8. Change of the position 1.00 (1.0) 1.7 14.7 32.4 45.8 5.3

9. Change of the bed linen 1.00 (1.0) 0.4 6.2 23 65.2 5.1

10. Emotional & psychological support 2.00 (1.0) 10.6 30.8 38.5 17.4 2.6

11. Necessary conversation 2.00 (2.0) 7.8 27.3 36.2 24.6 4.1

12. Information about therapies 2.00 (1.0) 4 21.0 36.9 33.5 4.7

13. Continence training (diapers) 1.00 (1.0) 6.8 14.6 22.7 31.1 24.7

14. Continence training (insert catheter) 1.00 (1.0) 1.2 4.7 12.5 52.5 29.1

15. Activating or rehabilitating care 2.00 (2.0) 5.2 23.2 40.1 22.8 8.8

16. Education and training 1.00 (1.0) 2 11.4 28.8 33.5 24.3

17. Preparation for discharge 2.00 (1.0) 1.9 14.2 39.8 35.4 8.8

18. Monitoring patients as described by physician 2.00 (1.0) 2.8 14.9 33.5 47.8 1

19. Monitoring patients as the nurse felt necessary 2.00 (1.0) 4.8 19 34.2 40.6 1.5

20. Monitoring of confused patients & use of restrains 1.00 (1.0) 4.9 18.7 19.9 32.9 23.6

21. Monitoring of confused patients & use of sedatives 2.00 (2.0) 5.1 19.9 25.6 27.5 21.8

22. Delay in measure because of a physician delay 2.0 (1.0) 3.8 18.6 28.9 31.9 16.8

23. Administration of medication, infusions 2.00 (1.0) 2.8 17.5 43.1 35.4 1.2

24. Change of wound dressings 1.00 (1.0) 0.8 9.6 28.1 54.7 6.8

25. Preparation for test and therapies 1.00 (1.0) 0.4 8.2 28.9 53.3 9.2

26. Keep patient waiting who rung 2.00 (1.0) 5.1 18.6 35.9 40 0.3

27. Adequate hand hygiene 1.00 (1.0) 2.8 10.5 34.9 51.5 0.2

28. Necessary disinfection measures 1.00 (1.0) 1.2 6.3 30 61.7 0.8

29. Studying care plans 2.00 (1.0) 7.6 16.7 32.3 42.6 0.7

30. Assessment of newly admitted patient 2.00 (2.0) 11.5 26.6 34.8 19.1 8

31. Set up care plans 2.00 (2.0) 12.3 23.7 34.2 21.8 8

32. Documentation & evaluation of the care 2.00 (2.0) 5.4 24.3 41 28 1.3

SD = standard deviation; IQR = interquartile range; BERNCA-R: range of the scale 0 = not required, 1 = never, 2 = rarely, 3 = sometimes, 4 = often.

Table 3

Summaries of the rationing predictor variables.

Variables (N = 1633) Nurse level Unit level Hospital level

Rationinga – mean (SD) 1.69 (0.558) 1.69 (0.223) 1.69 (0.141)

Quality of the nurse practice environmentb

(1) Nurse manager ability – mean (SD) 3.07 (0.635) 3.69 (0.395) 3.07 (0.297)

(2) Staff resources adequacy – mean (SD) 2.52 (0.685) 2.52 (0.428) 2.52 (0.309)

(3) Collegial nurses & physicians – mean (SD) 2.92 (0.524) 2.92 (0.282) 2.92 (0.178)

Workload

Patient-to-nurse ratio – mean (SD) 7.87 (5.139) 7.87 (2.029) 7.87 (1.472)

Number of patients needing support with all ADL – mean (SD) 4.29 (3.380) 4.29 (1.542) 4.29 (1.019)

Number of patients needing hourly/more frequently monitoring – mean (SD) 2.24 (2.313) 2.23 (0.830) 2.24 (0.551)

Safety climatec – mean (SD) 4.98 (1.338) 4.98 (0.544) 4.98 (0.386)a BERNCA-R: range of the scale 0 = not required, 1 = never, 2 = rarely, 3 = sometimes, 4 = often.b PES-NWI R: range 1 = strongly disagree, 2 = disagree, 3 = agree, 4 = strongly agree.c SOS: 1 = not at all, 2 = to a very limited extent, 3 = to a limited extent, 4 = to a moderate extent, 5 = to a considerable extent, 6 = to a great extent, 7 = a very

great extent; SD = standard deviation.

Please cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissacute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

patratiwatheassvaluestilow(TaNurhypin eadj

4. D

from

Tab

Sepa

of n

Va

St

Nu

Co

Pa

Nr

Nr

Sa

Nu

Nu

Sepa

lme

leve

on ra

4 = sb

4 = t

7 = a

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx 7

G Model

NS-2099; No. of Pages 10

Plac

ient-to-nurse ratio was significantly associated withoning. At the unit level, a higher patient-to-nurse ratios significantly associated with higher levels of rationing in

separate model (p-value 0.004) (Table 4), although thisociation was not significant in the adjusted model (p-

e 0.144) (Table 5). At the hospital level, a more favorablymated ‘‘patient safety climate’’ (H3) was associated wither levels of rationing in both the separate (p-value 0.003)

ble 4) and adjusted models (p-value 0.009) (Table 5).se experience and nurse education were not, asothesized (H4), significantly associated with rationingither the separate (p-value 0.162–0.855) (Table 4) or

usted models (p values 0.296–0.971).

iscussion

This study used data from the RN4CAST study, collected a quota sample of 35 Swiss acute care hospitals, to

describe the levels of implicit rationing in Swiss acute carehospitals and to explore, via multilevel modeling, theassociation between a set of potential correlates (selectedbased on the theoretical model) and implicit rationing ofnursing care. Comparable with the results of the RICHNursing study, the measured rationing levels acrosshospitals were rather low, although there was consider-able variability in rationing frequencies among the 32nursing tasks included in the BERNCA-R. Respondentsmost frequently rationed tasks in the areas of caring anddocumentation, and less frequently those related tomedical, technical, and therapeutic treatments. Theseresults are in line with the international evidence onrationing, prioritizing or missed nursing care, where mostrationing has also been observed in the areas of caring andsupport (Aiken et al., 2008; Morin and Leblanc, 2005;Rochefort and Clarke, 2010).

This suggests that nurses have developed an informalhierarchical system, which they use for prioritizing thenecessary nursing care, when resources are too scarce toprovide the required care to all patients. Tasks with a moredirect effect on patient’s health condition, which, ifomitted, are most likely to reduce negative consequenceson patient outcomes, e.g., monitoring or medicationadministration, seem to receive higher priority than thosewhose potential effects are less immediate or direct. Also,tasks for which the required amount of time cannot beexactly estimated, or where no possibility exists toshorten them by carrying out abbreviated versions,appear to be another factor driving prioritization. Forexample, the time needed to administer a medication or tochange a wound dressing can be more accuratelyestimated than the time needed to offer emotionalsupport. Or when changing the position of a patient withrestricted mobility/motility, faced with scarce resources/time, the nurse might decide to carry this task out with

le 4

rate multilevel models to identify significant predictors for rationing

ursing care.

riables B (estimates) Error t-Value p-Value

affing and resources adequacya (fixed estimates)

Nurse level 0.031 0.025 1.237 0.216

Unit level �0.181 0.055 �3.305 0.001

Hospital level �0.097 0.079 �1.227 0.210

rse manager abilitya (fixed estimates)

Nurse level 0.008 0.027 0.234 0.815

Unit level �0.064 0.066 1.707 0.088

Hospital level �0.121 0.082 0.394 0.679

llegial nurse physiciansa (fixed estimates)

Nurse level �0.036 0.030 �1.218 0.223

Unit level �0.013 0.079 �0.167 0.866

Hospital level �0.234 0.127 �1.911 0.062

tient-to-nurse ratio (fixed estimates)

Nurse level 0.004 0.003 1.526 0.127

Unit level 0.035 0.012 2.897 0.004

Hospital level �0.010 0.017 �0.600 0.543

. of patients needed support with all ADLs (fixed estimates)

Nurse level �0.001 0.004 �0.167 0.867

Unit level 0.025 0.015 1.677 0.095

Hospital level �0.007 0.025 0.296 0.768

. of patients needed � hourly monitoring (fixed estimates)

Nurse level 0.001 0.006 0.234 0.815

Unit level 0.048 0.028 1.707 0.088

Hospital level 0.017 0.042 0.394 0.679

fety climateb (fixed estimates)

Nurse level 0.003 0.011 0.243 0.808

Unit level 0.022 0.043 0.518 0.603

Hospital level �0.176 0.056 �3.164 0.003

rse experience (fixed estimates)

Nurse level 0.001 0.002 0.603 0.546

Unit level 0.006 0.006 1.023 0.306

Hospital level 0.008 0.012 0.649 0.502

rse education (fixed estimates)

Nurse level 0.050 0.050 1.001 0.317

Unit level �0.038 0.208 �0.185 0.855

Hospital level 0.355 0.257 1.380 0.162

rate multilevel models for each predictor performed with R package

4. The predictors were partitioned into three parts: nurse level, unit

l and hospital level, in order to investigate the effects of one predictor

ationing (outcome variable) at different levels.

PES-NWI R: range 1 = strongly disagree, 2 = disagree, 3 = agree,

trongly agree.

SOS: 1 = not at all, 2 = to a very limited extent, 3 = to a limited extent,

o a moderate extent, 5 = to a considerable extent, 6 = to a great extent,

very great extent.

Table 5

Three level model with significant predictors from separate models.

Variables B (estimates) Error t-Value p-Value

Staffing and resources adequacya (fixed estimates)

Nurse level 0.048 0.031 1.565 0.117

Unit level �0.161 0.081 �1.980 0.042Hospital level �0.049 0.135 �0.361 0.782

Patient-to-nurse ratio (fixed estimates)

Nurse level 0.005 0.004 1.514 0.127

Unit level 0.023 0.016 1.400 0.144

Hospital level �0.004 0.021 �0.178 0.847

Safety climateb (fixed estimates)

Nurse level 0.002 0.012 0.141 0.889

Unit level 0.072 0.043 1.683 0.090

Hospital level �0.190 0.073 �2.618 0.009

Multilevel regression analysis performed with R package lme4. Fixed and

random estimates of 3-level model were estimated. The random part of

the three-level model includes the unit level random intercept and

hospital level intercept. In the adjusted model was controlled for gender

(female), age, employment (fulltime) and hospital typology as possible

confounders.a PES-NWI R: range 1 = strongly disagree, 2 = disagree, 3 = agree,

4 = strongly agree.b SOS: 1 = not at all, 2 = to a very limited extent, 3 = to a limited extent,

4 = to a moderate extent, 5 = to a considerable extent, 6 = to a great extent,

7 = a very great extent.

ease cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx8

G Model

NS-2099; No. of Pages 10

only passive patient involvement, and not, as normally,with active involvement. However, this supposed prioritysystem requires further exploration, preferably using aqualitative study design.

Multilevel regression analysis was used to explore theassociation between the nine selected predictors andimplicit rationing of nursing care. Although internationalresearch often uses only the patient-to-nurse ratio inregression models, because of co-linearity between thisvariable and the NWI-R staffing and resources subscale,this study used both variables. This was based on the aim,as stated above, of evaluating the potential predictors forrationing. It is therefore important to evaluate which ofthese variables is the better predictor for rationing.Furthermore, in this study the co-linearity between thesetwo variables was within an acceptable range, whichjustifies this procedure.

As hypothesized, the unit level staff resource adequacy,a dimension of the nurse practice environment, wassignificantly associated with rationing (H1). Assuming thatnurses from the same unit will rate the quality of thepractice environment similarly, it is not surprising that,aggregated at the unit level, this variable was significantlyrelated with rationing. Because ‘staffing and resourceadequacy’ is a prerequisite for nurses to deliver adequatecare to their patients is this result, from a conceptual andclinical point of view not unexpected.

Neither of the other two studied nurse practiceenvironment dimensions – ‘nurse manager ability’, and‘collegial nurses & physicians’ was significantly associatedwith rationing in either the separate or the adjusted model,indicating that these two dimensions are less relevant forrationing. Because staffing and resource adequacy is aprerequisite for nurses to deliver adequate care to theirpatients, the relevance of this subscale as a predictor forrationing from a conceptual and clinical point of view is notsurprising. However, these results contradict earlierfindings, which suggest significant associations betweenboth dimensions and rationing (Rochefort and Clarke,2010; Schubert et al., 2007).

Of the three other selected organizational measures –the patient-to-nurse ratio, the number of patients whoneed support with all ADLs’, and ‘the number of patientswho need hourly or more frequently monitoring’ – (H2),only the unit-level patient-to-nurse ratio was signifi-cantly linked with rationing in the separate model, and nosuch association was visible in the adjusted model. Thisresult indicates that the nurse work environment ‘staffresource adequacy’ dimension, which includes items suchas ‘‘Enough staff to get the work done’’ or ‘‘Enoughregistered nurses on staff to provide quality patient care’’is a stronger predictor for rationing than the patient-to-nurse ratio. For future research it will be necessary to usemore refined measures of nurse staffing and workload.One possibility is nursing hours-per patient-day, whichreceived the highest consensus score in Van den Heedeet al.’s expert rating as an indicator of the number ofnursing staff necessary on a unit (Van den Heede et al.,2007). Using this variable Kalisch et al. found a significantassociation between higher numbers of hours per patientday and the incidence of missed nursing care, which is a

related concept to rationing of nursing care (Kalisch et al.,2011).

Confirming our hypothesized relationship betweenhospital safety climate/culture and rationing (H3), patientsafety climate, aggregated at the hospital level, was thesecond factor significantly associated with rationing.However, considering that the SOS reflects patient safetyclimate at the unit level it was somewhat surprising to finda significantly association at the hospital level instead. Onepossible explanation might be the small variability insafety climate levels among units within the samehospitals. However, as this is the first known case of asignificant association between safety climate and ration-ing, it requires further exploration.

The hypothesized associations between nurse experi-ence and nurse educational level and implicit rationing ofnursing care (H4) were not confirmed. This may haveresulted from the homogeneity of the nurse sample, i.e.,the fact that the overall proportion of nurses withacademic degrees was very likely too small (10%) for itseffect to be distinguishable. Furthermore, the number ofyears worked as a nurse, which we used as a proxy forexperience, may have been too unrefined in this context toshow an effect. And while these results support those ofKalisch et al., who found no significant associationbetween professional experience and missed nursing care(Kalisch et al., 2011), given that various studies andprojects indicate an association between professionalexperience/expertise and decision-making, clinical judg-ment and prioritization (Klein, 2008; Ruth-Sahd andHendy, 2005; Thompson et al., 2008), these associationsdemand further exploration. Future studies on the topicwould benefit from mixed method designs, the quantita-tive elements of which should include more refinedmeasures of expertise.

In fact, one of this study’s most significant limitationswas its cross-sectional design, which allowed no conclu-sions regarding causality. Furthermore the use of theRN4CAST’s pre-defined number of correlates of implicitrationing of nursing care ruled out exploring the full rangeof potential correlates suggested in the theoretical frame-work.

At the same time, several strengths are notable. In spiteof the limits of a cross-sectional design, for example, aquota sample of 35 small, midsize and larger Swiss acutecare hospitals represented Swiss acute care hospitals asaccurately as possible, thereby supporting the externalvalidity of its findings. Further, the results underscore theadequacy of the statistical approach used to explore eachpredictor’s effects on outcomes at three different levels.This permitted both the identification of significantpredictors and the definition of the degree to which eachwas linked with the outcomes of interest.

5. Conclusions and implications

This study’s results showed variability in the prevalenceof rationing among the 32 nursing tasks, indicating that,when faced with resource shortfalls, nurses prioritizenecessary nursing tasks in different ways. However, itappears that tasks related directly to medical, technical,

Please cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissacute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

andgen

linkmustafnurhostheindlevreq

praminlevthrpatlevobvthethebe

alsooutissuandlev

betinp201roleeduissuis rdesdivandthilyinessaltecarpat

Ack

theRNto

Shu

Uni201tion

eth

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx 9

G Model

NS-2099; No. of Pages 10

Plac

therapeutic treatments related to patient safety areerally accorded the highest priority.Of nine potential predictors, two were consistentlyed with rationing in both the separate and adjustedltilevel models. The first – unit level adequacy of nursefing resources – is a basic condition for safe, adequatesing care (Clarke and Donaldson, 2008). The second –pital level patient safety climate – was studied here for

first time in relation to rationing, with resultsicating a significant relationship between hospital-el patient safety climate and rationing. This associationuires further exploration.Regarding possible implications for clinical nursingctice, the identified rationing predictors can deter-e starting points for interventions: their prevalence

els can indicate when rationing exceeds predefinedesholds, i.e., where the risk of negative impacts onient and nurse outcomes rises to an unacceptableel (Schubert et al., 2008, 2009). With this in mind, oneious starting point would be to invest in improving

quality of the nurse work environment, particularly adequacy of nursing resources. As this could logicallyexpected to lead to reduced rationing levels, it would

reduce the risk – and expense – of negative patientcomes. Similarly, by increasing awareness of relatedes, investing in an institution’s patient safety climate

culture would also likely lead to reduced rationingels.Further, given the previously indicated relationshipsween rationing and patient safety, quality of care andatient mortality rates (Schubert et al., 2008, 2009,2), it is crucial to increase our understanding of thes patient safety climate, nurse experience andcation play in this context. Since knowledge of thesees cannot be captured with quantitative research, it

ecommended that future studies use mixed-methodigns to explore the interrelationships between theerse related factors, implicit rationing of nursing care

outcomes (as described in the framework used fors study). Finally, more knowledge of nurses’ under-

g decision-making and clinical judgment processes isential to develop strategies to guide them, when nornative exists, in prioritizing and rationing necessary

e in a way that minimizes negative impacts onients.

nowledgements

The authors thank the participating hospitals, nurses, contact persons in the hospitals, and the Swiss4CAST Stakeholders committee for their contributionthis research. The authors also acknowledge Chrisltis for his work editing the article.

Conflict of interest: No conflict of interest.Funding: This study was funded by the Europeanon’s Seventh Framework Programme (FP7/2007-3) under grant agreement 223468. For more informa-

on the RN4CAST project, please visit www.rn4cast.eu.Ethical approval: Approvals were obtained from 13 local

References

Aiken, L.H., Clarke, S.P., Sloane, D.M., Sochalski, J.A., Busse, R., Clarke, H.,Giovannetti, P., Hunt, J., Rafferty, A.M., Shamian, J., 2001. Nurses’reports on hospital care in five countries. Health Affairs 20 (3), 43–53.

Aiken, L.H., Clarke, S.P., Sloane, D.M., 2002a. Hospital staffing, organiza-tion, and quality of care: cross-national findings. International Journalfor Quality in Health Care 14 (1), 5–13.

Aiken, L.H., Clarke, S.P., Sloane, D.M., Lake, E.T., Cheney, T., 2008. Effects ofhospital care environment on patient mortality and nurse outcomes.Journal of Nursing Administration 38 (5), 223–229.

Aiken, L.H., Clarke, S.P., Sloane, D.M., Sochalski, J., Silber, J.H., 2002b.Hospital nurse staffing and patient mortality, nurse burnout, andjob dissatisfaction. JAMA 288 (16), 1987–1993.

Ausserhofer, D., Schubert, M., Engberg, S., Blegen, M., De Geest, S.,Schwendimann, R., 2012. Nurse-reported patient safety climate inSwiss hospitals: a descriptive–explorative substudy of the SwissRN4CAST study. Swiss Medical Weekly 142 , http://dx.doi.org/10.4414/smw.2012.13501.

BAG, Bundesamt fuer Gesundheit, 2008. Statistiken zur Krankenversi-cherung—Kennzahlen der Schweizer Spitaler 2005. In: Zahnd,D.B., Francis, A.B., Perret, A.O., Jobin, I.B. (Eds.), ElectronicBook. Vol. 1, Swiss Confederation. Departement des Inneren(EDI), Swiss Federal Office of Public Health, Bern, pp. 1–294,BAG Publikationsnr./no de publication OFSP/n di pubblicazioneUFSP: KUV 5.11 400dfi 60EXT1106, ISSN 1664-7270, Availablefrom: http://www.bundespublikationen.admin.ch.

Bruyneel, L., Van den Heede, K., Diya, L., Aiken, L., Sermeus, W., 2009.Predictive validity of the International Hospital Outcomes Studyquestionnaire: an RN4CAST pilot study. Journal of Nursing Scholar-ship 41 (2), 202–210.

Bucknall, T.K., 2000. Critical care nurses’ decision-making activities in thenatural clinical setting. Journal of Clinical Nursing 9 (1), 25–35.

Clarke, S.P., Donaldson, N.E., 2008. Nurse staffing and patient care qualityand safety. In: Hughes, R.G. (Ed.), Patient Safety and Quality: AnEvidence-Based Handbook for Nurses. Agency for HealthcareResearch and Quality, Rockville, MD, pp. 1–25.

Hancock, H.C., Easen, P.R., 2006. The decision-making processes of nurseswhen extubating patients following cardiac surgery: an ethnographicstudy. International Journal of Nursing Studies 43 (6), 693–705.

Hendry, C., Walker, A., 2004. Priority setting in clinical nursing practice:literature review. Journal of Advanced Nursing 47 (4), 427–436.

Jones, P.S., Lee, J.W., Phillips, L.R., Zhang, X.E., Jaceldo, K.B., 2001. Anadaptation of Brislin’s translation model for cross-cultural research.Nursing Research 50 (5), 300–304.

Kalisch, B.J., Tschannen, D., Lee, K.H., 2011. Do staffing levels predictmissed nursing care? International Journal for Quality in Health Care23 (3), 302–308.

Klein, G., 2008. Naturalistic decision making. Human Factors 50 (3),456–460.

Kuenzi, K., Schaer-Moser, M., 2002. The Labour Situation in Nursing in theCanton of Bern. (Die Arbeitssituation im Pflegebereich im KantonBern).Buro Bass Buro a&o, Bern, Switzerland, pp. 1–81 (in German).

Lake, E.T., 2002. Development of the practice environment scale of theNursing Work Index. Research in Nursing and Health 25 (3), 176–188.

Morin, D., Leblanc, N., 2005. Less money, less care: how nurses in long-term care allocate hours of needed care in a context of chronicshortage. International Journal of Nursing Practice 11 (5), 214–220.

Neuhaus, J.M., Kalbfleisch, J.D., 1998. Between- and within-cluster cov-ariate effects in the analysis of clustered data. Biometrics 54 (2),638–645.

Rochefort, C.M., Clarke, S.P., 2010. Nurses’ work environments, carerationing, job outcomes, and quality of care on neonatal units. Journalof Advanced Nursing 66 (10), 2213–2224.

Ruth-Sahd, L.A., Hendy, H.M., 2005. Predictors of novice nurses’ use ofintuition to guide patient care decisions. Journal of Nursing Education44 (10), 450–458.

Schubert, M., De Geest, S., 2003. Implizite Rationierung in der Pflege in derSchweiz—Eine Realitat? (Literature Analysis Unpublished). Instituteof Nursing Science, University of Basel, Basel, pp. 1–34.

Schubert, M., Clarke, S.P., Glass, T.R., Schaffert-Witvliet, B., De Geest, S.,2009. Identifying thresholds for relationships between impacts ofrationing of nursing care and nurse- and patient-reported outcomesin Swiss hospitals: a correlational study. International Journal ofNursing Studies 46 (7), 884–893.

Schubert, M., Glass, T.R., Clarke, S.P., Aiken, L.H., Schaffert-Witvliet, B.,Sloane, D.M., De Geest, S., 2008. Rationing of nursing care and itsrelationship to patient outcomes: the Swiss extension of the Inter-

national Hospital Outcomes Study. International Journal for Quality inHealth Care 20 (4), 227–237. ics committees responsible for the 35 study hospitals.

ease cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016

M. Schubert et al. / International Journal of Nursing Studies xxx (2012) xxx–xxx10

G Model

NS-2099; No. of Pages 10

Schubert, M., Glass, T.R., Clarke, S.P., Schaffert-Witvliet, B., De Geest, S.,2007. Validation of the Basel extent of rationing of nursing careinstrument. Nursing Research 56 (6), 416–424.

Schubert, M., Clarke, S.P., Aiken, L.H., De Geest, S., 2012. Associationsbetween rationing of nursing care and inpatient mortality in Swisshospitals. International Journal for Quality in Health Care, http://dx.doi.org/10.1093/intqhc/mzs009.

Sermeus, W., Aiken, L.H., Van den Heede, K., Rafferty, A.M., Griffiths, P.,Moreno-Casbas, M.T., Busse, R., Lindqvist, R., Scott, A.P., Bruyneel, L.,Brzostek, T., Kinnunen, J., Schubert, M., Schoonhoven, L., Zikos, D.,Rn4cast, R.C., 2011. Nurse forecasting in Europe (RN4CAST): rationale,design and methodology. BMC Nursing 10 (1), 6.

Squires, A., Aiken, L.H., van den Heede, K., Sermeus, W., Bruyneel, L.,Lindqvist, R., Schoonoven, L., Stromseng, I., Busse, R., Brozstek, T., Ensio,A., Moreno-Casbas, M., Rafferty, A.M., Schubert, M., Zikos, D., 2012. Asystematic survey instrument translation process for multi-country,

comparative health workforce studies. International Journal of NursingStudies, http://dx.doi.org/10.1016/j.ijnurstu.2012.02.015.

Thompson, C., Dalgleish, L., Bucknall, T., Estabrooks, C., Hutchinson, A.M.,Fraser, K., De Vos, R., Binnekade, J., Barrett, G., Saunders, J., 2008. Theeffects of time pressure and experience on nurses’ risk assessmentdecisions: a signal detection analysis. Nursing Research 57 (5), 302–311.

Van den Heede, K., Clarke, S.P., Sermeus, W., Vleugels, A., Aiken, L.H., 2007.International experts’ perspectives on the state of the nurse staffingand patient outcomes literature. Journal of Nursing Scholarship 39(4), 290–297.

Vogus, T.J., Sutcliffe, K.M., 2007. The Safety Organizing Scale: develop-ment and validation of a behavioral measure of safety culture inhospital nursing units. Medical Care 45 (1), 46–54.

West, E., Mays, N., Rafferty, A.M., Rowan, K., Sanderson, C., 2007. Nursingresources and patient outcomes in intensive care: a systematic reviewof the literature. International Journal of Nursing Studies 2007.

Please cite this article in press as: Schubert, M., et al., Levels and correlates of implicit rationing of nursing care in Swissacute care hospitals—A cross sectional study. Int. J. Nurs. Stud. (2012), http://dx.doi.org/10.1016/j.ijnurstu.2012.09.016