Characteristics of included studies
Thirty-one articles published between 1995 and 2021 were included in this review, reporting 29 unique studies. Thirteen articles were quantitative and 18 were qualitative. Figure 1 displays the flow diagram. Excluded full-text articles and individual reasons for exclusion are available in Additional file 2.
The main characteristics of the included articles are presented in Table 1. There was a wide variation in the characteristics of the studies, both in terms of A&F topic, duration of follow-up, and type of effects measured. The settings included hospitals, nursing homes, and community-based facilities.
Table 1
Characteristics of included quantitative and qualitative studies
Authors, year | Aim | Country | Design | Setting | Participants who received audit and feedback | MMAT overall score (%) |
Borgert 2016 (51) | To investigate the difference in effect on transfusion bundle compliance between two Audit and Feedback (A&F) strategies to implement the transfusion bundle | The Netherlands | Quasi-experimental comparative study design | 28-bed mixed medical-surgical Intensive Care Unit (ICU) of a university hospital | ICU nurses | 100 |
Cheater 2006 (52) | To evaluate the effectiveness of: 1) audit and feedback, 2) educational outreach, 3) audit and feedback in combination with educational outreach, and 4) printed educational materials alone (control) in promoting improvements in community nursing practice and patient outcomes | UK | Factorial design, cluster randomized controlled trial | Community health service - local family practices | Community nurses (district nurses and specialist health visitors) | 80 |
Choi 2014 (29) | To implement an evidence utilization project via an audit and feedback approach to improve cancer pain management | Korea | Implementation process (no design specified) | University affiliated, tertiary care hospital | Nurses in leadership positions and staff nurses | 40 |
Christina 2016 (53) | To explore the perceptions of nurses in an acute care setting on factors influencing the effectiveness of A&F. | Canada | Qualitative study | Teaching hospital, on two units specialising in internal medicine and haematology–oncology | Nurses who provide direct patient care and who have been exposed to A&F | 100 |
Clark 2018 (30) | To compare the impact of generalized feed back to staff with that of individualized, timely feedback. | USA | Pre-post study | Outpatient department of a large regional academic medical center serving | Registered nurses | 40 |
Cooper 2002 (36) | To identify the factors that determine the successful outcome of clinical audit programmes in the speciality of palliative care | UK | Action research | Palliative care setting | Patient services manager, clinical medical officer, social worker/counsellor, clinical nurse specialists, | 40 |
Dixon-Woods 2013 (54) | To evaluate an approach to improving patients’ safety and quality of care involving the secondary use of data from a hospital electronic prescribing and decision support system (ePDSS). | UK | Ethnography | Four clinical areas of an acute care hospital | Nursing staff, doctors | 100 |
Duncan 2000 (32) | To determine the effect of an individual performance feedback intervention provided to nurses on patient postoperative pain management outcomes | USA | Quasi-experimental, pretestposttest design | Orthopedic unit at a midsize acute care hospital | Registered nurses and Licensed practice nurses | 20 |
Giesbers 2016 (46) | To better understanding how feedback to nursing teams on quality measurements affects nurses’ well-being and performance | The Netherlands | Embedded case study | Surgical wards from one type of hospital, i.e. acute teaching-hospitals | Nurses | 80 |
Giesbers 2021 (43) | To provide a better understanding of nurses’ attributions about the reasons for providing them with feedback and the role of ward managers in creating a supportive feedback environment, in order to explain how providing feedback on quality measurements to nursing teams in a hospital setting is related to nurses’ engagement and burnout. | The Netherlands | Convergent mixed-methods, multiple case study design | Surgical wards from one type of hospital, i.e., general, teaching hospitals | Nurses | 80 |
Grando 2007 (38) | To examine further ways to improve the usefulness of QI feedback reports by explicating nursing home providers’ perceptions of Quality Improvement Feedback Interventions. | USA | Qualitative descriptive follow-up study | Nursing homes | Nursing staff, Director of Nursing, Administrator, Other staff | 40 |
Hermon 2015 (39) | To describe the use of our electronic tool to monitor and feedback process compliance in conjunction of introducing bespoke central line insertion packs to tackle catheter-related bloodstream infections | UK | Quasi-experimental quality improvement project | District general hospital. | Nursing and medical staff | 60 |
Hutchinson 2020 (40) | To develop and test the effect of an audit with feedback implementation strategy that was intended to increase the rate of voluntary medication error reporting by nurses | Australia | Quasi-experimental | Wards (two surgical and two medical) from a large, private, not-for-profit hospital. | Managers and full-/part-time nurses (both registered and enrolled nurses) | 80 |
Jeffs 2014 (44) | To explore the perceptions and experiences of front-line nurses and managers associated with the implementation of a unit-level dashboard in one teaching hospital. | Canada | Qualitative study | Hospital units | Staff nurses and managers | 80 |
Jeffs 2014 (59) | To explore the perceptions and experiences that nurses and managers had with their respective unit’s Care Utilising Evidence (CUE) dashboard. | Canada | Qualitative study | Units from a inner city acute care teaching hospital | Nurses and managers | 40 |
Jones 2020 (57) | To identify the ways in which different types of patient experience feedback act as a trigger or prompt for engagement in reflection in clinical practice | UK | Ethnographic study | Acute care units in three hospital trusts | Staff working in acute medical units (AMUs) and intensive care units (ICUs) | 100 |
Keizer 2020 (58) | To explore the healthcare workers’ needs and expectations for future antimicrobial resistance prevention measures audit and feedback. | The Netherlands | Qualitative study | High-risk antimicrobial resistance departments at a regional hospital | Physicians, residents and nurses | 80 |
Pedersen 2019 (45) | To describe and explore how first-line managers in a district psychiatric center experience using A&F cycle as basis for quality improvement in specialist mental health care. | Norway | Case study | District psychiatric center | First-line managers | 100 |
Peters 2018 (35) | To create a new automated audit-and-feedback process with the goal of improving post opioid-assessment documentation. | USA | Participatory action research | 97-bed community hospital | Staff nurses, charge nurses, nurse managers, and the nurse administrator | 40 |
Rantz 2001(31) | To determine if simply providing nursing facilities with comparative quality performance information and education about quality improvement would improve clinical practices and subsequently improve resident outcomes, or if a stronger intervention, expert clinical consultation with nursing facility staff, is needed. | USA | Three-group randomized experimental design | Nursing homes | Administrator and director of nursing; in many cases, a staff nurse | 20 |
Rhew 1999 (33) | To compare the effectiveness of three interventions designed to improve the pneumococcal vaccination rate | USA | Prospective controlled trial | General Medicine ambulatory clinic | Registered nurses and licensed practical nurses | 40 |
Righi 2020 (48) | To analyze the effectiveness of an audit and feedback programme for enhancing adherence to preventive practices and reducing facility-acquired pressure ulcers prevalence in a multicenter study of nursing homes followed for 2 years. | Switzerland | Prospective longitudinal | Nursing homes | Reference nurses, nursing home directors and nursing staff | 60 |
Ross 2016 (49) | To explore responses to the feedback of hospital performance on stroke quality of care measures among relevant physicians and other healthcare personnel at VHA hospitals | USA | Qualitative - interviews | Veterans Health Administration hospitals | Neurologists, internists, Emergency Department physicians, chiefs of staff, nurses, rehabilitation staff and quality management staff | 80 |
Sales 2014 (41) | To assess the effect of feedback reporting on rates of pain assessment, depression screening, and falls over time | Canada | Interrupted time series design with a comparison group | Long-term care nursing units in continuing care facilities | Facility administrators, nurse managers, and frontline direct care staff | 40 |
Schoonover-Shoffner 1995 (56) | To describe the meaning of performance feedback usefulness for individual feedback recipients receiving formal, objective performance feedback within a specified performance feedback situation | USA | Qualitative- naturistic approach | Tertiary-care metropolitan teaching hospital | Nurse managers | 100 |
Shaw 2019 (37) | To explore the attitudes of health professionals, health informaticians and information communication technology professionals to using data in electronic health records (eHRs) for performance feedback and professional development | Australia | Qualitative, in a co-design framework | Public hospitals | Nursing staff, doctors, information communication technology professionals, health informaticians, and allied health professionals | 40 |
Sykes 2021 (47) | To describe the content and application of hospital ward audit in order to identify potential enhancements to such audits | UK | Qualitative | Acute hospital provider organisations | Healthcare professionals - feedback is presented to ward managers | 80 |
Turner 2014 (55) | To evaluate the challenges of designing and implementing practice-based codification projects, including how they were influenced by professional interests and managerial priorities | UK | Qualitative | Two hospitals | Front-line clinicians | 80 |
Wald 2011 (42) | To audit patient-level postoperative catheter duration and measure the impact of its feedback to nursing staff on postoperative catheter duration and catheter-associated urinary tract infections rates | USA | Preobservational and postobservational study | Orthopedic and general surgery units | Registered nurses | 80 |
Wald 2014 (34) | To test the utility of feedback of indwelling urinary catheter (IUC) duration and Catheter-associated urinary tract infection (CAUTI) data on subsequent IUC duration using data obtained from a centralized, automated electronic surveillance strategy implemented at multiple hospitals participating in a CAUTI prevention collaborative | USA | Pragmatic, cluster, randomized, controlled trial | Nonintensive care units of Nurses Improving Care for Healthsystem Elders (NICHE) hospitals | Staff nurses | 40 |
Wilkinson 2000 (50) | To investigate reactions to the use of evidence-based cardiovascular and stroke performance indicators within one primary care group. To identify the range of perceptions towards the indicators and actions stimulated in response to them. | UK | Qualitative analysis of semi-structured interviews | Practices from a primary care group | General practitioners, practice managers, and practice nurses | 80 |
Quality appraisal
The overall methodological quality of the 13 included quantitative articles was low-moderate with scores ranging from 20 to 100. Six quantitative studies were of lower quality (29–34), mainly because aspects of the method lacked precision and were incomplete. Overall, the methodological quality of the 18 qualitative studies was good. Four studies were of lower quality (35–38), mostly because the interpretation of data was poorly supported by quotes and the qualitative approaches were not well defined. Overall, qualitative studies were of higher methodological quality than quantitative studies. Overall MMAT scores for each study are reported in Table 1. The breakdown of the scores by study and design is reported in Additional file 3.
Features of A&F interventions (Objective 1)
Most of the interventions’ features did not meet the recommendations by Brehault et al. (2016) (2) for optimizing the effectiveness of feedback. The quantitative studies overall provided much more comprehensive information on the features of A&F interventions than the qualitative studies. All intervention features are reported in a table format in Additional file 4.
What A&F information was delivered?
Most studies used team-level or unit-level measures (29, 33, 34, 39–47), facility-level measures (31, 48, 49) or practice-level measures (50). Six studies reported using individual-level data as feedback indicators (30, 32, 35, 51–53).
The feedback was about processes of care in twelve studies (29, 32, 35, 36, 45, 47–49, 51, 52, 54, 55), mostly on best practice compliance. Nine studies reported feedback on outcomes of care (30, 31, 33, 41, 42, 46, 50, 56, 57), while three others reported on feedback on both processes and outcomes of care (34, 39, 43). Other studies reported the general thematic of feedback without providing the specific content of feedback (37, 44, 53, 58).
Who was the A&F delivered to?
Of the studies that reported on how feedback was delivered, there was a nearly even distribution between interventions delivered in group (34, 39, 42, 43, 46, 47, 50, 55) (57) and those delivered individually (32, 33, 35, 41, 52, 53). Three studies (30, 51, 54) reported that feedback was delivered to nurses both individually and in groups. The Jeffs et al. (2014) (44, 59) studies reported that feedback was delivered variably in the participating units, sometimes individually and sometimes in group. Several qualitative studies did not specify this feedback feature (29, 31, 36, 37, 40, 48, 49, 56, 58).
Why was the A&F delivered?
Two studies (43, 50) reported that a clear explanation of the purpose, rationale, and characteristics of A&F was provided to participants before the intervention began. The other studies did not mention how or to what extent the concept of A&F was introduced to participants.
How was the A&F delivered?
The modes of A&F delivery were diverse. Most studies reported that feedback was given in writing, either in the form of a paper report (41, 56), email or mail (30, 33, 52), or a combination of posters displayed on the unit and email (40). In some studies, feedback was provided in both written and verbal form, although very little detail was provided on the delivery (32, 43, 45, 46, 48, 51). Three studies reported providing feedback only verbally (47, 50, 53). Two studies reported that feedback was provided in changing forms over the study period (44, 57).
Two studies reported including some suggestions on how nurses could make changes (52, 53). Four studies (30, 40, 43, 46) mentioned that the A&F specified clear targets that nurses were expected to meet. All remaining studies did not identify a specific behavior to be modified.
Three studies mentioned that nurses produced (45, 47) or were encouraged to produce (50) an action plan following the feedback. Among these, Sykes et al. (2021) (47) mentioned that an action plan was developed only if the performance on a specific norm was low to very low in two consecutive audits. No other study mentioned the preparation of an action plan as part of the A&F intervention. One quantitative (48) and one qualitive studies (43) mentioned that the feedback was delivered through discussion with team members, without specifying the manner in which the discussion was conducted or its content. As for the reflection phase of feedback, one qualitative study (45) reported that the team reflected on the feedback while another one (42) mentioned that nurses were asked to brainstorm on barriers following feedback. Overall, there was very little mention of the reflection and action planning phases across studies.
The use of educational sessions or materials was reported as a complementary strategy to A&F in 10 of 13 quantitative studies, four of which including at least one face-to-face workshop (29, 31, 39, 48, 52). Educational strategies in the other studies were mostly passive, including the provision of guidelines and written information (30, 40, 45, 50, 51) or video presentation (34, 42).
Other complementary strategies were reported sporadically and included the use of reminders (30, 33), accessibility to the materials needed for the targeted practice (39), creation of a stakeholder group to determine the methods and mechanisms for feedback (31, 40) and provision of training credits to participants (34). The use of theory to support intervention development was reported in only one study (40). One study reported using a methodological guideline for implementation (29).
Measured effects and their association with A&F features (Objectives 2 and 3)
The effects of A&F were measured on follow-up periods ranging from one to six months in six studies (29, 30, 33, 42, 51, 52). Other studies had longer periods of follow-up ranging from 12 to 24 months (31, 32, 34, 39–41, 48). Measured effects for all studies are presented in Additional file 5.
Studies with the highest methodological scores (39–42, 48, 51, 52) reported modest or no positive effects. Among these, only one study used a randomized controlled design (52). This study found no improvement at six-month follow-up on a series of nursing-sensitive process indicators following implementation of A&F (52). Most of the other studies used a quasi-experimental design (39–41, 51).
Two of these (39, 51) reported overall positive effects. Borgert, Binnekade (51) reported significant improvement in transfusion bundle compliance following an A&F intervention combining individual and group feedback. The group that received individual face-to-face performance feedback had more favorable results in the short term, but these results were not sustained after the intervention. Individual feedback was received verbally within 72 hours after transfusion, which was the most intensive A&F intervention in terms of repeating feedback. Results showed a significant improvement bundle compliance during the A&F period, but a significant decrease after it stopped. The authors further noted the lack of sustainability of this intervention in current practice. Similarly, Hermon, Pain (39) targeted compliance with central venous catheter care bundles. Feedback on process indicators was reported quarterly to nursing teams. The compliance with bundle was recorded through medical records in an added row daily following each nursing shift. This study reported significant improvements both in compliance with the measures and in infections rates 24 months from the start of the implementation.
Hutchinson et al. (2020) reported that A&F did not effectively influence nurses' self-reporting behaviors related to medication errors six months after implementation. Righi, Ourahmoune (48) reported little improvement in preventive measures and pressure ulcer prevalence, although many indicators remained unchanged after A&F at the two-year follow-up. Sales, Schalm (41) reported no improvement in process and outcome indicators for catheter-associated urinary tract infections at 13-month follow-up and for pain assessment, depression screening, and falls rates.
The remaining studies, which ranged in methodological quality from moderate to poor, reported generally better results, more so for process measures than for outcome measures (29–33). Among these, a randomized trial (31) compared the effects of combined A&F with education and A&F with clinical consultation in nursing homes. Scores of thirteen outcome indicators were measured at facility-level and reported to nurses in a written form. There was no significant differences between groups following the interventions. Rhew, Glassman (33) reported significantly improved assessment and vaccination rates following individual A&F and the use of reminders over a 3-month period. A&F on assessment and vaccination rates were reported to nurses twice monthly via email.
Studies in which feedback was delivered only in a written form, either on posters or by email, showed very little positive effects (40, 41, 52). Most studies (29–31, 34, 39, 40, 42, 45, 48, 51, 52) reported using education as an additional strategy to A&F. High-intensity educational strategies were not associated with improved outcomes. Feedback based on individual and team-level data generally produced more positive results (39, 51) compared to facility-level feedback (31, 48), although this was not consistent across all studies.
Qualitative findings on perceived A&F effects (Objective 4)
The first step in the analysis was to organize the data extracted from the qualitative studies according to the four components of Murray's model (22), separating the data for nurses from those for managers. The data were then organized according to the direction of the effect they reflected, either positive, neutral, or negative. From this core frame, themes that reflected the effects of the intervention were generated by identifying patterns within and across studies. Subthemes reflecting explanations underlying the effects were then generated. For example, in the cognitive participation component, data related to the indifference effect theme were organized into different subthemes, including lack of understanding of the data, lack of emotional significance, and lack of accuracy of the data. Overall, the qualitative data showed that nurses perceived several drawbacks in the way A&F is conducted, while recognizing the relevance of secondary use of data to improve care.
Coherence and signification - What does A&F mean to the nurses as a way to improve performance?
The main reactions expressed by nurses regarding the meaning given to A&F ranged from reluctance to willingness.
Reluctance to the delivery of A&F
On the one hand, many studies reported that nurses generally did not appreciate the way A&F was conducted (36, 43, 46, 47, 53). A&F interventions were perceived by some nurses as extra workload (36). Reluctance often stemmed from a fear of audit being used in a punitive way (37, 43, 46, 53, 54). Some nurses, based on previous experience with quality improvement interventions, felt that the A&F came with a hidden agenda, such as cost control or reputation maintenance, which was an irritant for them (43, 53). Nurses reported, however, that the negative perception of A&F could be mitigated if more explanation was provided about the purpose and process of A&F (44, 46, 53). The added value of A&F to nursing daily practice was rarely made explicit (44, 53).
Willingness to engage in A&F
On the other hand, some nurses were willing to participate in A&F interventions from the start. Those nurses perceived A&F as an opportunity for professional development and as being intrinsically motivating for improving the quality of care (37, 38, 44–46, 49).
Cognitive participation – To what extent does A&F engage nurses into improving performance?
Promoting awareness
Awareness was frequently evoked as an effect of A&F for both nurses and managers (43, 44, 46, 47, 56). Awareness was translated into paying attention, being alert, keeping in mind the measurements, but did not explicitly reflect an intention to act on the feedback.
Providing motivation to act on feedback
The higher degree of cognitive participation described in qualitative studies was a commitment to improving care following A&F. Nurses from one study reported that performance scores lower than expected stimulated and motivated the team to work on improvements (49). Many participants expressed that they valued the specificity of feedback on elements they could directly act on as nurses (35, 37, 49). One study (46) reported nurses expressing how the feedback acted as a motivation for addressing how to improve the results and an opportunity to commit in improving related aspects of care. Several nurses in another study (49) reported that A&F was a way to reflect expertise and current efforts to improve areas of care. Feedback acted as a validation of the team's ability to improve, which was perceived as engaging.
Leading to indifference
Less favorable responses from nurses in terms of engagement have been found in some studies. We observed some gradation in the intensity of cognitive responses, ranging from indifference to contempt.
Some studies reported rather neutral effects of A&F in nurses. Nurses expressed having a limited understanding of measures in a few studies (35, 50, 56), which limited their capacity to engage in A&F. Nurses in three studies expressed receiving too much information in feedback reports (44, 49, 58), which limited their ability to connect the dots and make sense of the indicators.
Nurse managers expressed that it was difficult to make individual nurses feel accountable for the results of team-level quality measures (46). Some nurses reported that they tended to feel accountable for the team performance when the scores were high but did not feel as responsible and questioned when performance was low (46, 53). This response occurred mostly when the nurse rated her own individual performance as satisfactory (50, 53).
Two studies reported that the indicators used in the feedback lacked clinical or emotional meaning (56, 57), which limited nurses’ response to the measures. According to nurse managers, providing situational referents, such as context and concrete examples shared by the team, could elicit greater nurse involvement (56). When indicators were incongruent with the priorities of nurses, they expressed being more likely to ignore feedback (47, 53, 56). Nurses also questioned whether the data captured the complex narratives of clinical practice (37, 53). Others expressed concern about how organizations focus primarily on negative feedback to promote improvement and leave out opportunities for reflection and change based on positive feedback (57).
Skepticism was a common response to A&F evoked by nurses (38, 50, 56). Some nurses and nurse managers expressed disbelief towards the accuracy of data, especially when performance scores were low (47). Concerns included the potential for inconsistent reporting in clinical system information, variation in measures across settings and misleading auto-reported data. Skepticism towards data more often led to indifference towards A&F (38, 50, 56).
Collective action - What do nurses do to make A&F function?
Discussing the feedback and potential actions
Several nurses expressed that A&F required and led to discussions within nursing teams and between colleagues (36, 44–46). Time allocated for discussion after A&F would help some nurses better understand the measures by allowing them to focus on the "story" behind the numbers. Reflecting as a team could give nurses a greater sense of clarity about what to do next (46).
Two studies (50, 55) in which feedback was reported in multidisciplinary meetings through discussion reported that physicians led the meetings while nurses were mostly silent, both in response to feedback and when potential actions were discussed. Wilkinson et al. (2000) (50) described nurses as being excluded from reflecting on actions related to improvements during feedback meetings with physicians in primary care, and Dixon-Woods (2013) (54) reported that nurses' ideas about organizational changes were not listened to during feedback meetings.
Attempting to act on the feedback
Frontline managers expressed more favorable opinions about the usefulness of A&F when they considered it to be an improvement tool in itself (38, 45, 49). Pedersen et al. (2019) noted that the managers interviewed in their study appeared to evade responsibility for the implementation of actions in practice. Ross (2016) noted that while participants expressed that A&F was very useful to them, how it led to change was not addressed.
Conversely, managers and nurses who expressed a sense of responsibility for implementing actions as a result of A&F noted the lack of support for how to achieve change (46, 47, 56) especially since A&F was considered as a process requiring endurance and long-term focus to sustain change (56). Some actions were perceived as being beyond the team's control, for instance in terms of a lack of material resources (47).
Nurses from one study in nursing homes reported that they felt isolated in their efforts to improve care following A&F, especially because there was a high degree of turnover and lack of stability within the teams (38). Other nurses described A&F as time-consuming, in a context where there was a perceived lack of time for direct patient care (36, 44).
Some nurses expressed that the feedback lacked directions on what was expected from them (46, 50). A study that targeted both physicians and nurses in primary care (50) noted that action plans tended to be shared informally among a few enthusiastic physicians and that the need for formal action plans was considered unnecessary, whereas changes are intuitive. On their part, the nurses felt that the action plan was necessary to provide a clear direction for action and to facilitate information sharing within the team. This study reported that nurses perceived a lack of team approach to change (50).
Misrepresenting the performance
Unintended actions following A&F were expressed in few interviews in the form of misrepresentation and actions contrary to what is expected. Managers expressed having misrepresented the actions taken following A&F by influencing the level of performance measured in the next audit (47, 54) or by completing an action plan knowing it would not be implemented. These managers expressed that writing a plan was a way to demonstrate a response to higher-level managers, either for fear of retribution or being targeted as a failing team or department, or because it was easier than carrying out actual actions (47). Sykes et al. (2021) reported that punitive feedback was replicated down a communication chain involving director of nursing, ward managers, and ward staff. Christina et al. (2016) (53) reported that some nurses ‘purposely’ avoided fulfilling audit criteria out of ‘spite’, due to frustration with the A&F process.
Reflexive monitoring – What benefits do nurses see in A&F?
Perceiving some positive benefits
The benefits of A&F were not widely addressed across the qualitative studies. Some nurse managers in one study reported that while A&F could improve team processes, it did not always translate into improved patient outcomes (45). Some nurses reported improvement in specific areas following A&F, for example in the accuracy of nursing assessment (36). Nurses from on study perceived that A&F could directly lead to improved patient outcomes, while recognizing not all their colleagues shared this view (46). Although this was not a consistent finding across studies, some nurses reported that improved data entry led to improved quality of patient care (46, 50).
Perceiving few tangible benefits
Several interviews with nurses and leaders emphasized the importance of A&F being perceived by recipients as beneficial in some capacity, which many did not. Perceived benefits for their own practice were rarely reported by nurses. One study (54) noted that nurses ended up feeling like they were being watched more closely and blamed than other professionals, especially physicians. These nurses were almost invariably called for account during feedback meetings, while physicians did not attend on a regular basis and therefore escaped rebuke. For the nurses, this meant not only that they were the only ones to bear the blame, but also that all the actors concerned by the problem were not involved, thus limiting the results of feedback on the quality of care. Based on their observations, Dixon et al. (2013) hypothesized that it was easier for managers to achieve accountability from nurses than physicians, possibly due to their lower professional status.
Several nurses expressed a dichotomous view of performance measures and patients’ welfare (43, 46), sometimes to the point of suggesting an incompatibility between the two concepts. Some participants' focus on improving documentation following A&F was seen as time-consuming, reducing their available time for patients, and having no real value on quality of care (46, 47, 54).
Concordance between qualitative and quantitative results (Objective 5)
Overall, there was concordance between measured and the perceived effects of A&F. Most of quantitative studies made no mention of how A&F was introduced to the nurses, only two studies reported using both verbal and written modes of feedback delivery and very few reported any discussion on feedback. Impacts of these characteristics were reflected in qualitative studies in which some nurses and managers mentioned having a limited understanding of the feedback content and its purpose. Both nurses and managers reported perceiving a limited impact of A&F on quality-of-care outcomes, which is consistent with the very few quantitative studies that reported positive results on outcome indicators.
Most of quantitative studies reported a high number of performance measures reported to nurses during A&F, which is consistent with some nurses’ perception that too much information was provided during feedback. Several nurses in qualitative studies reported not knowing how to act upon feedback, which is consistent with characteristics of quantitative studies, which hardly ever address target behaviours and actions related to feedback.
Individual feedback was generally associated with positive results in quantitative studies (30, 33, 51), although the effects were measured on very short term (no longer than on a 4-month period) in all three studies. One of those studies (51) reported a decrease in compliance at the end of the 4-month follow-up. Based on qualitative findings, individual A&F tended to be perceived as more threatening and generating punitive consequences (53, 54). While allowing for more nursing-specific measures, individual feedback was associated with a higher risk of being unfairly singled out, which contributed to reluctant and withdrawn responses to A&F. The combination of these findings suggests that individual feedback places more pressure on nurses to change their behavior, and thus may have short-term effects, but does not promote longer-term changes and may foster nurses' perception of A&F as a monitoring and control tool.
The two papers (31, 38) that reported both measured and perceived effects for the same trial did not show closely aligned results. While the A&F intervention resulted in no improvement in 13 indicators in a nursing home setting, nurses and managers both expressed having a high interest towards the intervention and being motivated following the delivery of feedback reports. This combination of findings suggests that favorable emotional reactions and cognitive participation with A&F may promote intention to change but are not sufficient to implement change.