Background Because patients are at the frontline of care where safety climate is closely tied to safety events, understanding patient perceptions of safety climate is crucial. We sought to develop and evaluate a parent-reported version of the Agency for Healthcare Research and Quality Hospital Survey on Patient Safety Culture and to relate parent-reported responses to parental need to watch over their child's care to ensure mistakes are not made.
Methods Parents (n=172) were surveyed about perceptions of hospital safety climate (14 items representing four domains—overall perceptions of safety, openness of staff and parent communication, and handoffs and transitions) and perceived need to watch over their child's care. Confirmatory factor analysis (CFA) was used to validate safety climate domain measures. Logistic regression was used to relate need to watch over care to safety climate domains.
Results CFA indices suggested good model fit for safety climate domains. Thirty-nine per cent of parents agreed or strongly agreed they needed to watch over care. In adjusted models, need to watch over care was significantly related to overall perceptions of safety (OR 0.20, 95% CI 0.11 to 0.37) and to handoffs and transitions (0.25, 0.14 to 0.46), but not to openness of staff (0.67, 0.40 to 1.12) or parent (0.83, 0.48 to 1.45) communication.
Conclusions Findings suggest parents can provide valuable data on specific safety climate domains. Opportunities exist to improve our safety climate's impact on parent burden to watch over their child's care, such as targeting overall perceptions of safety as well as handoffs and transitions.
- Patient safety
- Patient-centred care
Statistics from Altmetric.com
The crucial nature of understanding patient experiences with care has resulted in considerable progress in developing and applying these measures to improve outcomes such as adherence and satisfaction with care, including care for children.1–5 Patient perspectives on safety represent a potential complement to the existing measures of patient experiences and are suggested to be highly correlated with patient experiences of care.6 Patient perspectives on safety are also critically important because patients are at the frontline of care, where these perceptions may be most closely associated with important outcomes such as patient safety indicators and readmissions.7–11
Patients possess the knowledge and willingness to engage as partners in ensuring safe care to varying degrees, depending upon factors related to the patient personally (educational attainment), characteristics of the illness, the healthcare setting and the tasks involved.12–16 Numerous expert bodies recommend patient engagement in ensuring safe care,17–19 yet a recent Agency for Healthcare Research and Quality (AHRQ) commissioned report identified the lack of tools to assess patient views of safety as a crucial gap in current patient safety efforts.20 Safety climate, composed of the values, attitudes, perceptions, competencies, and patterns of behaviour within an organisation, can be a useful metric for understanding a hospital's proficiency with and commitment to safety.6 Understanding and assessing safety climate is a key strategy for improving patient safety.21 Although several survey-based tools exist to evaluate safety climate from the viewpoint of clinicians,6 ,22–27 to our knowledge, no instrument exists to evaluate safety climate from the viewpoint of patients or families.
Children are highly vulnerable to medical error,28 due in part to their dependence on adults’ communication and other behaviours to prevent error. For this reason, experts and governing bodies have suggested family engagement in care can improve safety for hospitalised children.29–32 Parents may well recognise their potential role in reducing error, with around two-thirds of parents of children hospitalised at a prominent US children's hospital feeling the need to watch over their child's care to ensure mistakes are not made.33 Yet, given the psychological distress experienced during a child's hospitalisation, perhaps the need for parents to watch over care is rather common and not necessarily reflective of parental perceptions of the hospital's safety climate.
To advance the field of paediatric patient safety, we sought to understand the potential value of parent reports about safety climate in children's hospitals. Specifically, we adapted and validated the AHRQ Hospital Survey on Patient Safety Culture as a tool for measuring parent perceptions of safety climate in children's hospitals, focusing on domains visible to parents—overall perceptions of safety, handoffs and transitions, and communication openness. We further examined how parent perceptions of safety climate were associated with their need to watch over their child's care.
During the period from October 2010 to May 2011, the family of each child hospitalised on our general paediatric hospitalist, pulmonology, haematology and oncology services across three hospital units was approached to participate in a study of parental perceptions of safety in our 60-bed, academic children's hospital. Our hospitalist service admits many children with acute concerns (eg, dehydration or respiratory distress) while the other services admit mostly children with ongoing or chronic illnesses (eg, cystic fibrosis, sickle cell disease or paediatric cancers). Children with stigmatising reasons for hospitalisation (eg, child neglect or mental health concerns) or whose parents were unavailable to consent (either absent or sleeping during our recruitment visits) were not eligible for participation. Our institution's review board (IRB) approved the study protocol.
Of 233 parents of eligible admissions in our continuous convenience sample, 194 agreed to learn about the study from our research team. Ultimately, parents of 172 of these admissions (87%) agreed to participate. When two parents were present, parents decided which parent would complete the consent process and surveys. Survey items included 14 items adapted from the AHRQ Hospital Survey on Patient Safety Culture6 to assess parental perceptions of safety climate in our children's hospital and a single, previously published item inquiring about the need to watch over the child's care to ensure mistakes are not made.33 Parent surveys also gathered data on parent (age, gender, education and race), child (age, gender and health status) and hospitalisation (reason for hospitalisation and length of stay) characteristics, based on known or theorised potential impacts on parent need to watch over care. Response options for reason for hospitalisation included a checklist of common reasons in lay terms (eg, breathing problem) as well as medical diagnoses (eg, asthma) and an ‘other’ response in which families could provide free text. In instances of multiple responses, all reasons for hospitalisation were retained and analysed. The free text responses were sorted into other checklist categories when appropriate or into new categories if needed.
Survey items were selected from the published literature and adapted by a multidisciplinary team, including health services researchers, clinicians, our hospital's patient safety officer, a health psychologist and experts in patient safety. The entire survey was pilot tested by eight parents purposively selected to represent various child ages, health status and exposure to the healthcare system. Parent feedback led to several changes, including adding definitions for some terms, addition of ‘not applicable’ on items that the family may not have experienced (eg, a transfer from one unit to another) and reordering of some response options for clarity and flow.
Items assessing parental perceptions of our children's hospital safety climate were adapted from the AHRQ Hospital Survey on Patient Safety Culture to reflect parent perspectives.6 We selected domains that focused on aspects of safety climate that are visible to parents, specifically overall perceptions of safety (four items), communication openness (three items reflecting staff communication openness), and handoffs and transitions (four items). To extend the domain of communication openness to represent openness of parent communication, we adapted the staff communication items to reflect a ‘parent communication’ domain (three items), assessing parent ability to communicate openly with our staff during the hospitalisation. Domain definitions and items are provided in table 1. Item response options were on a five-point Likert scale (1=strongly disagree, 5=strongly agree). Negatively worded items were reverse scored.
Parental perceptions of the need to watch over care to ensure mistakes do not happen was constructed as in the previously published study, with responses on a five-point Likert scale (1=strongly disagree, 5=strongly agree) and referred to hereafter as ‘parent need to watch over care’.33 In consideration of the response distributions and consistent with the published work, we dichotomised responses such that a response of 4 or 5 indicated a need to watch over the child's care. A sensitivity analysis found that including the neutral (neither agree nor disagree) response as reflecting agreement or disagreement with the need to watch over care did not substantially affect findings. Thus, we present findings from a conservative approach when this response was indicative of not needing to watch over care.
Means, SD and proportions were used to describe our participants, their perceptions of safety climate and their need to watch over the child's care. To ensure the measurement structure for the parent-reported safety climate domains was consistent with AHRQ's recommended measurement model for staff safety climate data,6 confirmatory factor analysis (CFA) with weighted least squares estimators was performed using parent survey responses collected within 24 h of admission. Standard model fit criteria were applied, including χ2/degrees of freedom (df), root mean square error of approximation (RMSEA), comparative fit index (CFI), Tucker–Lewis Index (TLI) and weighted root mean square residual (WRMR).34 ,35
Due to the strong positive correlations between the safety climate domains, four separate adjusted logistic regressions were performed, with each model relating parent need to watch over care to perceptions of one of the four safety climate domains, adjusting for parent, child and hospitalisation characteristics. To account for confounding by parent, child and hospitalisation characteristics and to construct the most parsimonious model, we used the accepted approach of including covariates associated in bivariate models at p<0.20 with our dependent variable, the need to watch over care.36 The results are summarised as adjusted OR and 95% CI. Analyses were performed using MPlus or STATA V.11.2. All tests of significance were at p<0.05 level and two tailed.
Parent, child and hospitalisation characteristics
Parents were predominantly mothers (84%) and non-minorities (83% white, non-Hispanic), with a wide range of educational attainment and an average age of 35 years (SD 7.75). On average, children were young (5.7 years (5.53)) and most were in good to excellent health (86%). The most common reasons for hospitalisation as reported by parents were breathing problems (30%), stomach or intestinal problems (20%) and fever (17%; table 2).
Parental perceptions of hospital safety climate
CFA on a single-factor model of hospital safety climate failed to demonstrate good model fit, using standard criteria. A four-factor model produced indices indicative of good model fit for assessment of the four safety climate domains (χ2/df=2.06, RMSEA=0.08, CFI=0.97, TLI=0.96, WRMR=0.86). τ-Equivalence testing supported equally weighting each indicator item, allowing construction of a summative scale for each of the four domains. Unstandardised and standardised factor loadings are available in table 1.
In general, parents viewed the safety climate in our institution positively. The most positively viewed safety climate domain was parent communication openness (mean 4.51 (0.58)). Overall perceptions of safety received a mean score of 4.15 (0.71), while staff communication openness and handoffs and transitions were rated similarly (3.97 (0.67) and 3.98 (0.72), respectively; table 3). Intercorrelations among the safety climate domains ranged from 0.40 to 0.83 (table 4).
Parent need to watch over care and safety climate
Thirty-nine per cent of parents agreed or strongly agreed they needed to watch over care to prevent mistakes. After adjustment for parent education and age, the child's gender and being hospitalised for breathing problems, parent need to watch over care was significantly inversely related to overall perceptions of safety (0.20, 95% CI 0.11 to 0.37) and to handoffs and transitions (0.25, 95% CI 0.14 to 0.46). Thus, on average, when parents’ overall perceptions of safety were one point higher, the odds of needing to watch over care decreased 80%. Similarly, when parent perceptions of our handoffs and transitions were one point higher, the odds of needing to watch over care decreased 75%. Parent need to watch over care was not significantly associated with staff (0.67, 95% CI 0.40 to 1.12) or parent (0.83, 95% CI 0.48 to 1.45) communication openness (table 5).
In all four adjusted models, parent need to watch over care was significantly associated with the child having been hospitalised for breathing problems compared with all other reasons for hospitalisation. Also, in adjusted models of the association of parent need to watch over care with handoffs and transitions, parent education was significantly associated with the parent need to watch over care. Compared with parents with some college education, parents with more education (3.01, 95% CI 1.31 to 6.90) were significantly more likely to need to watch over care.
Our work advances the field of paediatric patient safety by adapting and validating a measurement model for parent views of safety climate in a children's hospital. We find that the previously validated models for staff survey data fit parent data well for domains that are visible to parents—overall perceptions of safety, handoffs and transitions, and communication openness. Such a measure could be useful in taking a patient-centred approach to patient safety and for ensuring safety climate is assessed at the frontline of care, where it most often correlates with safety events.7–11 Further, two of the parent-reported safety climate domains are significantly associated with parent perceptions about needing to watch over care. Specifically, perceived need to watch over care was significantly associated with overall perceptions of safety and with handoffs and transitions. Our findings suggest potential targets to improve hospital safety climate and ultimately perhaps reduce parent burden for watching over care.
Beyond the recognised value of safety climate measures arising from the frontline of care, parent-reported safety climate measures for children's hospitals are needed for several reasons. Although hospitalised children disproportionately suffer medical errors, existing measures of patient safety are quite limited in their utility for the paediatric inpatient population and often focus on rare, high-impact medical errors such as central line infections, more commonly seen in intensive care unit settings. Such measures are poorly suited to smaller institutions and for the general paediatric population, most of whom do not have central lines and in which it may be years before a single event such as a hospital-acquired central line infection is documented. Further, being flexible is a key element of organisations with a strong safety culture.21 This flexibility includes deferring to expertise over rank and parents are assuredly the leading experts on their children's care.
Two other advantages of a parent-reported measure of safety climate arise from the characteristics of the data provided by parents. First, parent response rates to the survey were substantially higher than those of staff surveys.10 While this may be because of the research nature of this work, parents may be highly motivated to report about the safety of care provided to their child. Second, parent data are more easily analysable by the hospital unit. In general, hospitalised children and their parents remain on the same unit throughout a hospital stay, barring a deterioration that requires transfer to intensive care services. This is in contrast to the experience of physicians or other staff who may work across multiple units. Understanding safety climate at the unit or service level has been recommended, as safety climates are ‘local’ and analysing at the institution level may mask opportunities for improving safety.37 ,38
As noted in the AHRQ-commissioned review, healthcare system employees and patients or parents may view safety differently, but both views have value.39 In our work, parents rated handoffs and transitions, in addition to overall perceptions of safety, considerably higher than AHRQ national data from hospital staff, which are similar to data from our institution.40 Also, the minimum value for staff communication openness was quite high, perhaps reflecting either parents’ reluctance to report negatively about this aspect of care or that staff are careful about interactions that are visible to families. Further, parental experiences with specific aspects of hospital care (eg, the policies and procedures that could prevent errors) may be limited, affecting their ability to report on these. Future studies might survey staff and patients/parents to examine similarities and differences in perceptions of safety climate. In addition, such studies could examine the responsiveness to change for parental reports of safety climate to understand whether these could be used to assess the impact of intervention to improve paediatric patient safety.
The associations found between parent need to watch over care and safety climate domains suggest potential targets to improve hospital safety climate and ultimately perhaps reduce parent burden for watching over care. Specifically, the likelihood that a parent felt the need to watch over care was significantly greater when parents held less positive views of our children's hospital's overall safety climate or our handoffs and transitions. To address overall perceptions of safety, the focus might include optimising procedures and systems for preventing mistakes and making these procedures and systems visible to parents. Also, this domain's items tap into parent need to see a commitment to safety that exceeds our drive to get the work done. Thus, focusing on ensuring families feel safe may be especially crucial when our hospital is busiest. For handoffs and transitions, the focus likely resides in demonstrating that staff are imparting important information during these critical times. For example, performing nursing shift changes at the bedside with the parent may reduce parent need to watch over care.41 ,42 Since this study, our institution has implemented this type of nursing handoff at the bedside. Also, being explicit with parents about anticipated changes in healthcare team members (physicians or nurses) and perhaps doing these handoffs with the parent present could reduce parent need to watch over care.
Through our adjusted analyses, we were also able to consider how characteristics of the parent, the child and the hospitalisation impacted parent perceptions of safety. Specifically, having a child hospitalised for a breathing problem was associated with greater parent need to watch over care compared with being hospitalised for any other reason. In addition, parent education significantly influenced parents’ beliefs about whether they needed to watch over care to ensure mistakes were not made. Taken together, these findings support the critical need for considering a broad array of patient socio-demographics, hospitalisation characteristics and institutional factors as covariates in paediatric patient safety research, especially when using parent-reported data.
As with all observational studies, certain limitations should be considered. First, as a single institution study, the findings may not generalise to children's hospitals broadly. The sample size may have underpowered some comparisons. Also, variability in the communication openness domain scores was less than for the other two domains, perhaps limiting our ability to detect significant associations. One possible explanation for the limited variability in communication openness measures may be the relatively large number of highly educated parents in our study. In addition, our population has relatively few minority families, so we may under represent the views of those who are more likely to experience adverse events43 and also less likely to report these.44 Future work through a multi-institutional collaborative could examine these types of questions and safety climate broadly.
Due to IRB requirements, we have limited data on the parents who refused to learn about the study or did not consent. Given that absent or sleeping parents constituted our most common reason for study exclusion, our findings could be biased towards more negative views of safety climate and greater need to watch over care (ie, parents who felt the need to watch over care were more likely to remain with their child). However, our recruitment rate for eligible parents was high, exceeding that of the prior similar study.33 Because of the need to limit the survey to a reasonable numbers of items, we also lack data on concepts such as general beliefs about the safety of the healthcare system broadly and on prior experiences with medical error, which could influence views about safety.
In summary, we find that parent responses to our adapted version of the AHRQ Hospital Survey on Patient Safety Culture display good model fit with the domains described in the original instrument. Understanding patient experiences with care provides additional information for improving care, as these experiences are linked to outcomes such as satisfaction.45 Efforts are underway to understand parent experiences with their hospitalised children's healthcare, adapting the Hospital Consumer Assessment of Healthcare Providers and Systems for use with this population, although this tool does not focus on hospital safety.46 Having a tool to assess parent perceptions of safety climate in a children's hospital has numerous advantages and can act as a supplement to staff data. In addition, we were able to identify specific opportunities to improve our safety climate in an effort to help parents feel less burden for watching over their child's care to ensure mistakes are not made.
The authors gratefully acknowledge the generosity of the families who participated in the study.
Contributors EDC conceived and designed the study, secured funding, oversaw all data collection and analyses, drafted and revised the manuscript, and approved the final version for submission. PC secured funding, assisted with data collection, revised the manuscript, and approved the final version for submission. KWH assisted with data collection, created the dataset, revised the manuscript, and approved the final version for submission. VPR performed analyses, revised the manuscript, and approved the final version for submission. RLB performed or oversaw analyses, revised the manuscript, and approved the final version for submission. PJR oversaw analyses, revised the manuscript, and approved the final version for submission. LLD secured funding, assisted with data collection, revised the manuscript, and approved the final version for submission. MMK assisted with data collection, revised the manuscript, and approved the final version for submission. LAB assisted with data collection, revised the manuscript, and approved the final version for submission.
Funding This work was funded through an Agency for Healthcare Research and Quality R18 to EDC. The project described was also supported by the Clinical and Translational Science Award programme, previously through the National Center for Research Resources grant 1UL1RR025011, and now by the National Center for Advancing Translational Sciences grant 9U54TR000021. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or other funders. The funders had no role in designing, conducting or reporting the results of this study.
Competing interests None.
Ethics approval UW Madison Health Sciences Minimal Risk Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Due to IRB requirements, no additional data are available.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.