International Journal of Nursing Studies 52 (2015) 403–420
Contents lists available at ScienceDirect
International Journal of Nursing Studies journal homepage: www.elsevier.com/ijns
Review
Factors contributing to Registered Nurse medication administration error: A narrative review Angela M. Parry *, K. Louise Barriball, Alison E. While King’s College London, Florence Nightingale School of Nursing and Midwifery, UK
A R T I C L E I N F O
A B S T R A C T
Article history: Received 28 June 2013 Received in revised form 2 July 2014 Accepted 10 July 2014
Objective: To explore the factors contributing to Registered Nurse medication administration error behaviour. Design: A narrative review. Data sources: Electronic databases (Cochrane, CINAHL, MEDLINE, BNI, EmBase, and PsycINFO) were searched from 1 January 1999 to 31 December 2012 in the English language. 1127 papers were identified and 26 papers were included in the review. Data were extracted by one reviewer and checked by a second reviewer. Review methods: A thematic analysis and narrative synthesis of the factors contributing to Registered Nurses’ medication administration behaviour. Bandura’s (1986) theory of reciprocal determinism was used as an organising framework. This theory proposes that there is a reciprocal interplay between the environment, the person and their behaviour. Medication administration error is an outcome of RN behaviour. Results: The 26 papers reported studies conducted in 4 continents across 11 countries predominantly in North America and Europe, with one multi-national study incorporating 27 countries. Within both the environment and person domain of the reciprocal determinism framework, a number of factors emerged as influencing Registered Nurse medication administration error behaviour. Within the environment domain, two key themes of clinical workload and work setting emerged, and within the person domain the Registered Nurses’ characteristics and their lived experience of work emerged as themes. Overall, greater attention has been given to the contribution of the environment domain rather than the person domain as contributing to error, with the literature viewing an error as an event rather than the outcome of behaviour. Conclusion: The interplay between factors that influence behaviour were poorly accounted for within the selected studies. It is proposed that a shift away from error as an event to a focus on the relationships between the person, the environment and Registered Nurse medication administration behaviour is needed to better understand medication administration error. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Behaviour Contributing factors Medication error Medication administration error Nurses Patient safety Narrative review
What is already known about the topic? * Corresponding author at: Florence Nightingale School of Nursing & Midwifery, King’s College London, James Clerk Maxwell Building, 57 Waterloo Road, London, UK. Tel.: +44 207 848 3579. E-mail address: [email protected] (A.M. Parry). http://dx.doi.org/10.1016/j.ijnurstu.2014.07.003 0020-7489/ß 2014 Elsevier Ltd. All rights reserved.
Medication error is a major risk to patient safety attracting global attention. Medication administration is a key nursing role and a core component of the medication management process.
404
A.M. Parry et al. / International Journal of Nursing Studies 52 (2015) 403–420
Medication administration error shows no sign of abating despite considerable empirical, policy and professional attention. What this paper adds The framework of reciprocal determinism is applied in this review to comprehensively explore the factors contributing to Registered Nurse medication administration behaviour. Within the current medication administration error literature there is a greater focus on the organisation of healthcare to the neglect of Registered Nurse behaviour. Medication administration error could be better understood by a clearer focus on the inter-relationships between the individual person, the environment and the behaviour that results in error. 1. Introduction and background Patient safety is a global healthcare concern (WHO, 2004; The Health Foundation, 2013). Registered Nurses (RNs) as direct providers of care have an integral role in keeping patients safe. Medication error, including medication administration error, is the most frequent cause of preventable morbidity and mortality in hospitals (Adams and Koch, 2010; Cousins et al., 2011). This has therefore been identified as a priority patient safety issue (IoM, 2007) and a central concern for the nursing profession (ICN, 2002; Choo et al., 2010). Recent evidence, however, shows that medication administration error rates remain unacceptably high (AHRQ, 2013; EMA, 2013; NRLS, 2014). Literature reviews conducted a decade apart reported that similar factors were consistently identified as contributing to RN medication administration error (O’Shea, 1999; Brady et al., 2009). These reviews focused upon the individual and system factors to explain the occurrence of medication administration error events and presented a uni-directional relationship between the contributing factor, or factors, and the error event as proposed by Reason’s (1990) explanatory framework. Whilst recognizing that individual clinicians work in complex error-prone circumstances, this approach failed to acknowledge that professional registration demands that RNs are accountable practitioners and thus answerable for their acts and omissions (ICN, 2012; NMC, 2008). Registered Nurse medication administration behaviour incorporates both the actions that RNs engage in to keep patients safe during medication administration and those that lead to unintended error or violation of recommended behaviours to prevent medication administration errors. Medication administration safety is most usually understood and measured against the five behavioural rights of medication administration (Department of Health, 2004; WHO, 2011). Medication administration behaviour is therefore distinct from the medication administration error itself, with error being the product of the RN behaviour. Exploration of medication administration error as an outcome of human behaviour rather than as an individual event may offer new insights regarding the contributors to medication administration error.
2. Theoretical model Bandura’s (1986) theory of reciprocal determinism acknowledged the mutual relationship that exists between the three domains of environment, person and behaviour. Within this theory behaviour is determined by the individual through cognitive processes and other personal attributes, and by the bi-directional influence of their environment with people as both products and producers of their environment (Bandura, 1986). This creates interactions resulting in a triadic dynamic interplay within the three domains, with different sources of influence acknowledged to be of different strengths and arising at different times. Bandura’s (1986) theory has underpinned some studies and its three domains were reflected in the portrayal of safety culture by Cooper (2000). The utility of the theory in extending understanding of RN behaviour has however not previously been explored This theory provides a framework to understand the contributing factors underlying RN medication administration behaviours which result in medication administration error events; medication administration error events being determined as the product of the reciprocal interplay of the three domains. 3. Method 3.1. Aims and objectives This review aimed to examine international research relating to the contributing factors to RNs’ behaviour that result in a medication administration error event. The research question guiding the review was: What factors contribute to RN behaviour in medication administration error? 3.2. Search strategy Six electronic databases (Cochrane, MEDLINE, CINAHL, BNI, Embase, PsycINFO) were systematically searched for papers using facets derived from the following subject headings and key words: Registered Nurse, contributory factors, and medication (administration) error The search was limited to English language journals published between January 1999, an influential year in the patient safety movement (IoM, 1999; Department of Health, 2000), and December 2012. No relevant reviews were located within the Cochrane database. Due to the volume of literature obtained and the varied limits available within the chosen databases, advice was sought from an information subject specialist regarding applying further limits to the search. The addition of the term ‘research’ plus its derivatives as a fourth component to the search was recommended and applied. 3.3. Treatment of results A search of the databases identified 1127 results, with most being located through CINAHL. Following the removal of duplicates the remaining items were assessed
A.M. Parry et al. / International Journal of Nursing Studies 52 (2015) 403–420
for significance in three phases: (1) titles were screened for relevance, with removal of those not obviously relating to RN medication administration error and contributory factors. (2) Abstracts for each of the remaining papers were read and papers were either included or excluded on the basis of their principal focus, primary research papers concerned with factors contributing to RN medication administration error or safe medication administration practice. (3) Papers were then assessed against the inclusion criteria for a focus on studies undertaken in countries with developed healthcare systems, in acute adult hospital inpatient settings. Studies that focused on the outcomes to RNs involved in a medication administration error or interventions in response to the reporting of a RN
medication administration error were excluded together with studies about medication administration error in relation to specific health issues or specific medications. Studies were excluded if they related to mental health clients or primary care settings. Simulation studies and studies hosted in a single ward or unit were also excluded. The remaining studies were appraised using the STROBE reporting guidelines for observational research (ISPM, 2007) and studies rated as low were removed, leaving 26 studies (STROBE rating: high = 11; moderate = 15). A hand search of reference lists from the selected papers provided no additional studies. Data were extracted by the lead author and the process checked by a second reviewer. The literature identification process is summarised in Fig. 1.
Screening
Identification
Records identified through data base search (n=1127)
Records after duplicates with removed: (n=881)
Records excluded (n=246)
Records screened by title (n=881)
Records excluded as title showed that study was not primarily focused on RN medication administration error (n=138)
Abstracts assessed for eligibility (n=743). No additional records identified through other sources
Abstracts excluded as not primary research focused on factors contributing to RN medication administration error (n=483)
Included
Eligibility
Full text assessed for eligibility (n=260)
405
Studies excluded (n= 228) with rationale: Studies excluded were not: focused on acute Adult in-patient units (incl. simulation) (n= 102) Studies excluded were: focused on outcomes for RNs not factors contributing to MAE (n= 19); focused on error in relation to a specific medication or an MAE associated with a specific medical condition (n= 89); hosted on a single ward/unit (n= 18)
Studies retained for detailed appraisal (n=32)
Removal of studies based on Strobe rating as low (n=6)
Studies retained for inclusion in literature review (n=26)
No further studies identified by hand search of reference lists of selected studies
Fig. 1. Literature identification process.
A.M. Parry et al. / International Journal of Nursing Studies 52 (2015) 403–420
406
3.4. Data analysis The variety of instruments and reported data precluded a meta-analysis. Therefore a thematic analysis and narrative synthesis was performed to examine the literature using Bandura’s (1986) theory of reciprocal determinism as an organising framework. A narrative synthesis is defined by the textual approach that provides a trustworthy story of the findings from the selected literature (Popay et al., 2006). Specifically the literature was examined to identify: What factors in the workplace environment contribute to RN medication administration behaviour when an error occurs? What RN person factors contribute to RN medication administration behaviour when an error occurs? Identification of the key themes in the narrative synthesis was an iterative process. Selected studies were read and key themes and sub-themes extracted in relation to Bandura’s (1986) theory of reciprocal determinism. This was undertaken primarily by one reviewer, and checked by a second reviewer with disagreements discussed until agreement was achieved. 4. Findings
concern surrounding the topic. Studies were published between 2003 and 2012 with a peak in publications between 2006 and 2007. All studies were quantitative in design and gathered data primarily through self-report (n = 21), with five employing direct observation. Using medication administration error as the outcome of behaviour, the domains of the environment and person (Bandura, 1986) have been used to synthesise the findings from the selected studies (see Fig. 2). Table 1 summarises the design of the selected studies, and their findings. Within each domain, key themes and sub-themes provide further clarity as to the prominence of contributory factors within the current evidence base. Table 2 summarises the spread of reported factors contributing to RN medication administration behaviour which influence the occurrence of error. 4.2. Domain: Environment Two themes emerged within the environment domain: the clinical activity in the healthcare setting, and the type and disposition of the working environment. Two subthemes were identified. Theme 1 (Clinical workload). Within the overarching theme of clinical activity, four subthemes were identified as contributing to RN medication administration error, namely: RN staffing; RN workload; interruptions and distractions; and patient characteristics.
4.1. Overview of papers
RN staffing
The 26 selected studies had worldwide origins (n = 13 United States/Canada; n = 4 Europe; n = 4 Australia; n = 4 Asia; n = 1 multi-national) emphasising the global
Nearly half of the studies (n = 12) explored whether there was an association between medication administration error and RN staffing. Staffing was measured in
BEHAVIOUR Error Violation
Triadic reciprocality
PERSON RN characteristics Experience and expertise Demography RNs lived experience of work Fatigue and shift pattern Quality of working life
ENVIRONMENT Clinical activity RN staffing RN workload Interruptions and distractions Patient characteristics The work setting Hospital / unit structure Leadership and organisation Teamwork and communication Safety measures
Fig. 2. Findings from selected studies as seen through the theory of reciprocal determinism (Bandura, 1986).
Table 1 Summary of selected studies. Design
Sample
Data collection method
Al-Kandari and Thomas (2008) Kuwait
Cross-sectional survey
Convenience sample: 780 RNs (RR = 95%) in wards in 5 hospitals
Structured self-report questionnaire
Berdot et al. (2012) France
Non participant structured observation Longitudinal survey
Convenience sample: 28 RNs/ 108 patients (RR = 100%) 4 wards in 1 hospital Convenience sample: 279 medical/surgical nursing units (RR = 97.5%) in 146 randomly selected hospitals
Newly developed structured observation tool Structured self-report questionnaires
Fasolino and Snyder (2012) USA
Cross-sectional survey
Convenience sample: 161/ 163RNs (RR = 81%/65.7%) in 11 units in 1 hospital
Newly developed structured self-report questionnaire
Fogarty and McKeon (2006) Australia
Cross-sectional survey
Convenience sample: 176 RNs (RR = 63%) in 11 hospitals
Structured self-report questionnaires
Hofmann and Mark (2006) USA
Longitudinal survey
Structured self-report questionnaires
Manojlovich and DeCicco (2007) USA
Cross-sectional survey
Convenience sample: 1127 RNs (RR not reported) 81 medical surgical ward/units in 42 hospitals Convenience sample: 462 RNs (RR = 53%) 25 ITUs in 8 hospitals
Mark et al. (2003) USA
Cross-sectional survey
Chang and Mark (2009) USA
Convenience sample: 1682 RNs (RR = 73.8%). 1326 patients in 124 medical/surgical wards (RR = 91%) in 64 acute care hospitals (RR = 94%)
Structured self-report questionnaires
Structured self-report questionnaires
Factors contributing to RN medication administration behaviour when errors occur
Strobe rating
Environment
Person
!" MAE: Emergency events (r = 0.102 p = 0.05); "RN workload (r = 0.3, p = 0.001); RNs undertaking non-nursing task (clerical work) (r = 0.14, p = 0.001). !" MAE: Injectable medication (OR vs. oral 6.89 [95% CI: 4.06, 11.70] p < 0.001). !" MAE: "medication-related support services ! "non = severe errors (NSE) (p < 0.01).
–
Moderate
–
Moderate
!" MAE: "RN experience ! " non-severe error (p < 0.01). ! # MAE: "RN expertise ! #NSE (p < 0.01); "RN education ! #severe errors (p < 0.01) optimized at 54% graduate RNs on unit. !" MAE: #RN age (r = 0.13, p = 0.03); #RN experience (r = 0.16, p = 0.001).
High
!" MAE: Effective team membership (r = 0.19, p < 01). Positive perceptions of practice environment (r = 0.15, p < 0.01). !" MAE: # organisational climate (violations: r = 0.25 p = 0.01; errors r = 0.15, p = NR).
!# MAE: "Unit safety climate (r = 1.51,
Contents lists available at ScienceDirect
International Journal of Nursing Studies journal homepage: www.elsevier.com/ijns
Review
Factors contributing to Registered Nurse medication administration error: A narrative review Angela M. Parry *, K. Louise Barriball, Alison E. While King’s College London, Florence Nightingale School of Nursing and Midwifery, UK
A R T I C L E I N F O
A B S T R A C T
Article history: Received 28 June 2013 Received in revised form 2 July 2014 Accepted 10 July 2014
Objective: To explore the factors contributing to Registered Nurse medication administration error behaviour. Design: A narrative review. Data sources: Electronic databases (Cochrane, CINAHL, MEDLINE, BNI, EmBase, and PsycINFO) were searched from 1 January 1999 to 31 December 2012 in the English language. 1127 papers were identified and 26 papers were included in the review. Data were extracted by one reviewer and checked by a second reviewer. Review methods: A thematic analysis and narrative synthesis of the factors contributing to Registered Nurses’ medication administration behaviour. Bandura’s (1986) theory of reciprocal determinism was used as an organising framework. This theory proposes that there is a reciprocal interplay between the environment, the person and their behaviour. Medication administration error is an outcome of RN behaviour. Results: The 26 papers reported studies conducted in 4 continents across 11 countries predominantly in North America and Europe, with one multi-national study incorporating 27 countries. Within both the environment and person domain of the reciprocal determinism framework, a number of factors emerged as influencing Registered Nurse medication administration error behaviour. Within the environment domain, two key themes of clinical workload and work setting emerged, and within the person domain the Registered Nurses’ characteristics and their lived experience of work emerged as themes. Overall, greater attention has been given to the contribution of the environment domain rather than the person domain as contributing to error, with the literature viewing an error as an event rather than the outcome of behaviour. Conclusion: The interplay between factors that influence behaviour were poorly accounted for within the selected studies. It is proposed that a shift away from error as an event to a focus on the relationships between the person, the environment and Registered Nurse medication administration behaviour is needed to better understand medication administration error. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Behaviour Contributing factors Medication error Medication administration error Nurses Patient safety Narrative review
What is already known about the topic? * Corresponding author at: Florence Nightingale School of Nursing & Midwifery, King’s College London, James Clerk Maxwell Building, 57 Waterloo Road, London, UK. Tel.: +44 207 848 3579. E-mail address: [email protected] (A.M. Parry). http://dx.doi.org/10.1016/j.ijnurstu.2014.07.003 0020-7489/ß 2014 Elsevier Ltd. All rights reserved.
Medication error is a major risk to patient safety attracting global attention. Medication administration is a key nursing role and a core component of the medication management process.
404
A.M. Parry et al. / International Journal of Nursing Studies 52 (2015) 403–420
Medication administration error shows no sign of abating despite considerable empirical, policy and professional attention. What this paper adds The framework of reciprocal determinism is applied in this review to comprehensively explore the factors contributing to Registered Nurse medication administration behaviour. Within the current medication administration error literature there is a greater focus on the organisation of healthcare to the neglect of Registered Nurse behaviour. Medication administration error could be better understood by a clearer focus on the inter-relationships between the individual person, the environment and the behaviour that results in error. 1. Introduction and background Patient safety is a global healthcare concern (WHO, 2004; The Health Foundation, 2013). Registered Nurses (RNs) as direct providers of care have an integral role in keeping patients safe. Medication error, including medication administration error, is the most frequent cause of preventable morbidity and mortality in hospitals (Adams and Koch, 2010; Cousins et al., 2011). This has therefore been identified as a priority patient safety issue (IoM, 2007) and a central concern for the nursing profession (ICN, 2002; Choo et al., 2010). Recent evidence, however, shows that medication administration error rates remain unacceptably high (AHRQ, 2013; EMA, 2013; NRLS, 2014). Literature reviews conducted a decade apart reported that similar factors were consistently identified as contributing to RN medication administration error (O’Shea, 1999; Brady et al., 2009). These reviews focused upon the individual and system factors to explain the occurrence of medication administration error events and presented a uni-directional relationship between the contributing factor, or factors, and the error event as proposed by Reason’s (1990) explanatory framework. Whilst recognizing that individual clinicians work in complex error-prone circumstances, this approach failed to acknowledge that professional registration demands that RNs are accountable practitioners and thus answerable for their acts and omissions (ICN, 2012; NMC, 2008). Registered Nurse medication administration behaviour incorporates both the actions that RNs engage in to keep patients safe during medication administration and those that lead to unintended error or violation of recommended behaviours to prevent medication administration errors. Medication administration safety is most usually understood and measured against the five behavioural rights of medication administration (Department of Health, 2004; WHO, 2011). Medication administration behaviour is therefore distinct from the medication administration error itself, with error being the product of the RN behaviour. Exploration of medication administration error as an outcome of human behaviour rather than as an individual event may offer new insights regarding the contributors to medication administration error.
2. Theoretical model Bandura’s (1986) theory of reciprocal determinism acknowledged the mutual relationship that exists between the three domains of environment, person and behaviour. Within this theory behaviour is determined by the individual through cognitive processes and other personal attributes, and by the bi-directional influence of their environment with people as both products and producers of their environment (Bandura, 1986). This creates interactions resulting in a triadic dynamic interplay within the three domains, with different sources of influence acknowledged to be of different strengths and arising at different times. Bandura’s (1986) theory has underpinned some studies and its three domains were reflected in the portrayal of safety culture by Cooper (2000). The utility of the theory in extending understanding of RN behaviour has however not previously been explored This theory provides a framework to understand the contributing factors underlying RN medication administration behaviours which result in medication administration error events; medication administration error events being determined as the product of the reciprocal interplay of the three domains. 3. Method 3.1. Aims and objectives This review aimed to examine international research relating to the contributing factors to RNs’ behaviour that result in a medication administration error event. The research question guiding the review was: What factors contribute to RN behaviour in medication administration error? 3.2. Search strategy Six electronic databases (Cochrane, MEDLINE, CINAHL, BNI, Embase, PsycINFO) were systematically searched for papers using facets derived from the following subject headings and key words: Registered Nurse, contributory factors, and medication (administration) error The search was limited to English language journals published between January 1999, an influential year in the patient safety movement (IoM, 1999; Department of Health, 2000), and December 2012. No relevant reviews were located within the Cochrane database. Due to the volume of literature obtained and the varied limits available within the chosen databases, advice was sought from an information subject specialist regarding applying further limits to the search. The addition of the term ‘research’ plus its derivatives as a fourth component to the search was recommended and applied. 3.3. Treatment of results A search of the databases identified 1127 results, with most being located through CINAHL. Following the removal of duplicates the remaining items were assessed
A.M. Parry et al. / International Journal of Nursing Studies 52 (2015) 403–420
for significance in three phases: (1) titles were screened for relevance, with removal of those not obviously relating to RN medication administration error and contributory factors. (2) Abstracts for each of the remaining papers were read and papers were either included or excluded on the basis of their principal focus, primary research papers concerned with factors contributing to RN medication administration error or safe medication administration practice. (3) Papers were then assessed against the inclusion criteria for a focus on studies undertaken in countries with developed healthcare systems, in acute adult hospital inpatient settings. Studies that focused on the outcomes to RNs involved in a medication administration error or interventions in response to the reporting of a RN
medication administration error were excluded together with studies about medication administration error in relation to specific health issues or specific medications. Studies were excluded if they related to mental health clients or primary care settings. Simulation studies and studies hosted in a single ward or unit were also excluded. The remaining studies were appraised using the STROBE reporting guidelines for observational research (ISPM, 2007) and studies rated as low were removed, leaving 26 studies (STROBE rating: high = 11; moderate = 15). A hand search of reference lists from the selected papers provided no additional studies. Data were extracted by the lead author and the process checked by a second reviewer. The literature identification process is summarised in Fig. 1.
Screening
Identification
Records identified through data base search (n=1127)
Records after duplicates with removed: (n=881)
Records excluded (n=246)
Records screened by title (n=881)
Records excluded as title showed that study was not primarily focused on RN medication administration error (n=138)
Abstracts assessed for eligibility (n=743). No additional records identified through other sources
Abstracts excluded as not primary research focused on factors contributing to RN medication administration error (n=483)
Included
Eligibility
Full text assessed for eligibility (n=260)
405
Studies excluded (n= 228) with rationale: Studies excluded were not: focused on acute Adult in-patient units (incl. simulation) (n= 102) Studies excluded were: focused on outcomes for RNs not factors contributing to MAE (n= 19); focused on error in relation to a specific medication or an MAE associated with a specific medical condition (n= 89); hosted on a single ward/unit (n= 18)
Studies retained for detailed appraisal (n=32)
Removal of studies based on Strobe rating as low (n=6)
Studies retained for inclusion in literature review (n=26)
No further studies identified by hand search of reference lists of selected studies
Fig. 1. Literature identification process.
A.M. Parry et al. / International Journal of Nursing Studies 52 (2015) 403–420
406
3.4. Data analysis The variety of instruments and reported data precluded a meta-analysis. Therefore a thematic analysis and narrative synthesis was performed to examine the literature using Bandura’s (1986) theory of reciprocal determinism as an organising framework. A narrative synthesis is defined by the textual approach that provides a trustworthy story of the findings from the selected literature (Popay et al., 2006). Specifically the literature was examined to identify: What factors in the workplace environment contribute to RN medication administration behaviour when an error occurs? What RN person factors contribute to RN medication administration behaviour when an error occurs? Identification of the key themes in the narrative synthesis was an iterative process. Selected studies were read and key themes and sub-themes extracted in relation to Bandura’s (1986) theory of reciprocal determinism. This was undertaken primarily by one reviewer, and checked by a second reviewer with disagreements discussed until agreement was achieved. 4. Findings
concern surrounding the topic. Studies were published between 2003 and 2012 with a peak in publications between 2006 and 2007. All studies were quantitative in design and gathered data primarily through self-report (n = 21), with five employing direct observation. Using medication administration error as the outcome of behaviour, the domains of the environment and person (Bandura, 1986) have been used to synthesise the findings from the selected studies (see Fig. 2). Table 1 summarises the design of the selected studies, and their findings. Within each domain, key themes and sub-themes provide further clarity as to the prominence of contributory factors within the current evidence base. Table 2 summarises the spread of reported factors contributing to RN medication administration behaviour which influence the occurrence of error. 4.2. Domain: Environment Two themes emerged within the environment domain: the clinical activity in the healthcare setting, and the type and disposition of the working environment. Two subthemes were identified. Theme 1 (Clinical workload). Within the overarching theme of clinical activity, four subthemes were identified as contributing to RN medication administration error, namely: RN staffing; RN workload; interruptions and distractions; and patient characteristics.
4.1. Overview of papers
RN staffing
The 26 selected studies had worldwide origins (n = 13 United States/Canada; n = 4 Europe; n = 4 Australia; n = 4 Asia; n = 1 multi-national) emphasising the global
Nearly half of the studies (n = 12) explored whether there was an association between medication administration error and RN staffing. Staffing was measured in
BEHAVIOUR Error Violation
Triadic reciprocality
PERSON RN characteristics Experience and expertise Demography RNs lived experience of work Fatigue and shift pattern Quality of working life
ENVIRONMENT Clinical activity RN staffing RN workload Interruptions and distractions Patient characteristics The work setting Hospital / unit structure Leadership and organisation Teamwork and communication Safety measures
Fig. 2. Findings from selected studies as seen through the theory of reciprocal determinism (Bandura, 1986).
Table 1 Summary of selected studies. Design
Sample
Data collection method
Al-Kandari and Thomas (2008) Kuwait
Cross-sectional survey
Convenience sample: 780 RNs (RR = 95%) in wards in 5 hospitals
Structured self-report questionnaire
Berdot et al. (2012) France
Non participant structured observation Longitudinal survey
Convenience sample: 28 RNs/ 108 patients (RR = 100%) 4 wards in 1 hospital Convenience sample: 279 medical/surgical nursing units (RR = 97.5%) in 146 randomly selected hospitals
Newly developed structured observation tool Structured self-report questionnaires
Fasolino and Snyder (2012) USA
Cross-sectional survey
Convenience sample: 161/ 163RNs (RR = 81%/65.7%) in 11 units in 1 hospital
Newly developed structured self-report questionnaire
Fogarty and McKeon (2006) Australia
Cross-sectional survey
Convenience sample: 176 RNs (RR = 63%) in 11 hospitals
Structured self-report questionnaires
Hofmann and Mark (2006) USA
Longitudinal survey
Structured self-report questionnaires
Manojlovich and DeCicco (2007) USA
Cross-sectional survey
Convenience sample: 1127 RNs (RR not reported) 81 medical surgical ward/units in 42 hospitals Convenience sample: 462 RNs (RR = 53%) 25 ITUs in 8 hospitals
Mark et al. (2003) USA
Cross-sectional survey
Chang and Mark (2009) USA
Convenience sample: 1682 RNs (RR = 73.8%). 1326 patients in 124 medical/surgical wards (RR = 91%) in 64 acute care hospitals (RR = 94%)
Structured self-report questionnaires
Structured self-report questionnaires
Factors contributing to RN medication administration behaviour when errors occur
Strobe rating
Environment
Person
!" MAE: Emergency events (r = 0.102 p = 0.05); "RN workload (r = 0.3, p = 0.001); RNs undertaking non-nursing task (clerical work) (r = 0.14, p = 0.001). !" MAE: Injectable medication (OR vs. oral 6.89 [95% CI: 4.06, 11.70] p < 0.001). !" MAE: "medication-related support services ! "non = severe errors (NSE) (p < 0.01).
–
Moderate
–
Moderate
!" MAE: "RN experience ! " non-severe error (p < 0.01). ! # MAE: "RN expertise ! #NSE (p < 0.01); "RN education ! #severe errors (p < 0.01) optimized at 54% graduate RNs on unit. !" MAE: #RN age (r = 0.13, p = 0.03); #RN experience (r = 0.16, p = 0.001).
High
!" MAE: Effective team membership (r = 0.19, p < 01). Positive perceptions of practice environment (r = 0.15, p < 0.01). !" MAE: # organisational climate (violations: r = 0.25 p = 0.01; errors r = 0.15, p = NR).
!# MAE: "Unit safety climate (r = 1.51,
