Article Text
PDF
Abstract
Background Antenatal corticosteroids (ANCS) reduce complications of preterm birth; however, not all eligible women receive them. Many hospitals and providers do not have the right processes and conditions to enable ANCS administration with high reliability. The objective of this study was to understand conditions that enable delivery of ANCS with high reliability among hospitals participating in an Ohio Perinatal Quality Collaborative (OPQC) ANCS project.
Methods We conducted focus groups and semistructured interviews with members of the OPQC project team (n=27) and other care providers (n=70) using a purposeful sample of 6 sites involved in the OPQC ANCS project. Participants including nurses (n=57), attending obstetricians (n=17), physician trainees (n=21) and certified nurse midwives (n=2) were asked to reflect on their experiences and to identify factors contributing to optimal use of ANCS. Focus groups and interviews were transcribed verbatim and were analysed by a multidisciplinary team using an iterative approach that combined inductive and deductive methods to identify and categorise themes.
Results Six major themes supporting reliable implementation of ANCS at these hospitals emerged including: (1) presence of a high reliability culture, (2) processes that emphasise high reliability, (3) timely and efficient administration process, (4) multiple disciplines are involved, (5) evidence of benefit supports ANCS use and (6) benefit is recognised at all levels of the care team.
Conclusions Our findings identify the key processes and supports needed to ensure delivery of ASCS with high reliability and are reinforced by implementation and reliability science. They are useful for foundation of the successful implementation of other evidence-based practices at high levels of reliability.
- Healthcare quality improvement
- Qualitative research
- Implementation science
- Obstetrics and gynecology
- Human factors
Statistics from Altmetric.com
- Healthcare quality improvement
- Qualitative research
- Implementation science
- Obstetrics and gynecology
- Human factors
Background
Preterm birth is a leading cause of neonatal morbidity and mortality in the USA.1 ,2 Appropriate use of therapeutic agents that reduce the significant and costly morbidities associated with early delivery is critical in addressing the burden of prematurity. Use of antenatal corticosteroids (ANCS) in women expected to deliver a preterm infant of 24–34 weeks gestation is supported by the highest level of evidence and its efficacy has been well established for over 40 years.3–5 While rates of ANCS administration now exceed 80%, it has been suggested that missing 20% of eligible women still indicates opportunity for improvement.6–8 In these cases, reliability principles and implementation science have important roles in improving the application of evidence in practice to even higher levels.
Reliability science provides insight into the conditions necessary for performance at high levels over long periods of time.9–11 The principles of reliability science can be used to design processes and organisational structures that increase the likelihood that a system will consistently perform its intended function without over-reliance on (often) faulty human capabilities.12 Studies examining the application of evidence in practice have shown that, in many cases, appropriate care is delivered only 50–60% of the time.13 ,14 Reliability science has the potential to help increase the consistency with which appropriate evidence is implemented in practice because it provides support for a range of design characteristics that can be integrated into healthcare systems to improve reliability. These design characteristics include standardisation, checklists, feedback, reminders, decision aids, process redundancy and default actions.10
The field of implementation science has also provided a lens with which to understand (and improve) the application of evidence into practice. Implementation science involves the study of methods to promote the systematic use of evidence in routine care.15 Implementation science theories suggest that effective implementation must consider the characteristics of the intervention, the context in which the intervention is implemented (inner and outer settings), characteristics of the individuals involved and the process of implementation.16 Factors such as evidence strength and quality, cultural norms, leadership engagement, implementation readiness, social networks among individual adopters, and implementation processes that emphasise deferring decision making to front-line teams and including feedback on progress, have all been identified as key facilitators of implementation of evidence in practice.16 ,17
To promote more systematic and highly reliable administration of ANCS, the Ohio Perinatal Quality Collaborative (OPQC) began a quality improvement (QI) effort to increase rates of ANCS administration to ≥90% across the state in October 2011. Although OPQC anticipated seeing a more significant opportunity for improvement, data showed rates of ANCS administration across the state were greater than 90% early in the project.18 This provided an opportunity to learn more about the conditions that enable delivery of evidence-based practices such as ANCS with high reliability in this subset of hospitals. We hypothesised that lessons learned about the processes and conditions for the implementation of evidence with high reliability would also be applicable to other cases in healthcare.
Methods
Qualitative inquiry was selected as the optimal methodology to obtain a deeper understanding of the conditions that enable delivery of ANCS with high reliability among OPQC hospitals. The qualitative study was embedded within an overall process evaluation of the OPQC ANCS project.
Sampling of sites
Nineteen of the 20 charter member OPQC hospitals participated in the ANCS project. The OPQC charter hospitals constitute a small proportion (18.7%) of the maternity hospitals in the state, but account for 47% of the births. The charter hospitals are clustered in the six major metropolitan areas of Ohio and care for the highest-risk women.18 ,19
Site visits were conducted at 6 of the 19 participating OPQC sites (32%) 8–10 months after the intervention began. In order to learn as much as possible, a purposive sample was designed to select hospitals which varied on the following characteristics: rates of ANCS administration (table 1); engagement by the project team, ranging from very high to very low, as determined subjectively by OPQC QI staff; self-reported approach to treating women with ANCS as either aggressive versus conservative as determined by early OPQC efforts to have hospitals adopt a specific ANCS treatment algorithm; and facility characteristics including, location, size and academic affiliation. Rates of ANCS administration across all OPQC sites ranged from 86% to 98% with approximately a third of participating hospitals having a rate <90%. There was similar variation in rates across the six hospitals visited as part of this study.
Characteristics of sample hospitals (October 2011 to June 2013)
Sampling of participants at sites
During the site visit, we conducted interviews and focus groups with members of the OPQC ANCS QI team which included an operational team leader (key contact), an obstetrical physician lead and one to five other individuals with mutual accountability for the project. In an effort to elicit opinions and experiences from staff not connected to the QI team, we also recruited a minimum of five obstetrical nurses and five obstetrical providers (including trainees and nurse midwives) at each site. Staff who were not part of the QI team were recruited by the hospital key contact using posters and emails.
We conducted one focus group with QI team members and one interview with the physician lead at each site. These sessions were designed to last approximately 45–60 min and were meant to explore roles and experiences with the OPQC project as well as impressions about ANCS use. The QI team was interviewed without the physician lead so that the team members could more freely comment on the role of the physician lead in the OPQC project and the role of physicians, more generally, in rates of ANCS use.
Interviews of obstetrical nurses and providers not connected to the QI team were conducted using either an individual interview (approximately 15 min in length) or profession-specific focus group format (approximately 25 min in length). The choice to use an interview or focus group format was made based on site preference. Interviews and focus groups with obstetrical staff were shorter than the corresponding sessions with members of the OPQC ANCS QI team as they were more narrowly focused on awareness of the OPQC project and impressions about ANCS use.
A total of 97 individuals participated in 38 individual interviews, 15 small focus groups (2–6 individuals), and 2 larger focus groups (>6 individuals) at the six hospitals. Information on the characteristics of these participants is provided in table 2.
Characteristics of participants
Semistructured interviews and focus groups
Semistructured open-ended question guides for focus groups and individual interviews were developed collaboratively by the multidisciplinary investigator team. Trained members of the research team (HCK, LG and SNS) conducted the interviews and focus groups. We used open-ended questions to initiate discussion which allowed respondents to elaborate in their own words and bring up topics not included on the guides. Follow-up probes were used to expand and clarify responses. Topics on the question guides used for QI team members and the obstetrical physician lead were designed to address topics of interest for the process evaluation and this qualitative study. Topics included changes made as part of the OPQC project (if any), aspects of local context that were barriers or facilitators of general QI, and opinions on factors leading to high rates of ANCS administration. Topics on the question guides used for obstetrical nurses and providers focused on awareness of OPQC, opinions on factors leading to high rates of ANCS administration and briefly on hospital support for QI. All interviews and focus groups were conducted and audio recorded in a room that ensured privacy. Recordings were transcribed verbatim by a trained medical transcriptionist and checked for accuracy.
Data analysis
The analysis team consisted of a neonatologist and QI researcher (HCK) and a qualitative researcher (SNS), as well as a trained research assistant (CC). We initiated data analysis with a deductive approach using the question guide as an initial template for the coding framework. As we began to read and review the transcripts, we added codes and definitions to this framework in an inductive manner to reflect participants’ ideas and concepts related to ANCS use and the process of conducting the ANCS QI project.20 ,21 The analysis team met regularly in person after coding small sets of transcripts, refined the coding framework iteratively and applied it to the remaining transcripts. All coding decisions were discussed, differences resolved by consensus and codes recorded in an electronic database (Microsoft Excel). During subsequent rounds of analysis, the team systematically examined the electronic record of statements assigned to each code, collapsed some coding categories and discussed the interrelationships between themes.20 ,22 By using analysts with different professional disciplines, training and perspectives, discussion proceeded in a collaborative, reflective fashion, helping to ensure that no one researcher's biases predominated.22 ,23 We primarily drew on data generated discussing opinions on factors leading to high rates of ANCS for this qualitative study.
Findings
Our analyses revealed six major themes that explain hospitals’ high rates of ANCS administration including: (1) presence of a high reliability culture, (2) processes that emphasise high reliability, (3) timely and efficient administration process, (4) multiple disciplines are involved, (5) evidence of benefit supports ANCS use and (6) benefit is recognised at all levels of the care team. We discuss each theme along with illustrative quotes selected to represent all of the sites that were visited, different provider types, and QI team members and general hospital staff.
High reliability culture
Participants expressed attitudes that underscored the presence of key elements of a High Reliability Culture within these OPQC hospitals including a preoccupation with missed cases, empowerment of front line nursing staff and teamwork.
Staff spoke about having a ‘preoccupation with failure’—a constant awareness of cases where ANCS administration was missed. Participants described a range of formal mechanisms to identify failures and to learn from them, including posting data about missed cases, reviewing missed cases at meetings, conferences or huddles, and providing data feedback about missed cases. Most formal practices, specifically reviewing failures and posting data about missed cases, were introduced as part of the OPQC project and were more prominently featured at two sites (Hospitals C and E with rates of ANCS administration of 97% and 96%, respectively).You have a sense of what you're doing, then, you really have to look at what you are actually doing to see where the problems are…I don't think we would have tumbled on to some of the things that we did without looking at where we were missing—QI Team Physician 4 (Hospital C)
When you see the posters, it kind of reminds you of… the factories that have… the safety record, you know like—it has been so many days since the last accident…I guess I have never thought of it as a safety practice but it's safeguarding a baby—Nurse 7 (Hospital E)
We're trying to…, drill down, into those few cases that we're missing to see why, what happened… So reviewing them, bringing them up, talking to the providers who were taking care of the patient, really give them the information, you missed a case and this is what we think you could have done differently—QI Team Physician 2 (Hospital E)
Although formal strategies to identify failures were most evident at a small number of sites, informal conditions supporting a ‘preoccupation with failure’ were featured more widely. Staff (those directly involved with OPQC and not) spoke about their desire to eliminate failures and voiced disappointment when an eligible woman delivers without having received steroids.If we haven't gotten someone steroids and their delivering before 34 weeks, it resonates. It doesn't make me feel very good about it.—Provider 7 (Hospital A)
I think in the instances where I've been involved in a team that has failed to get them [ANCS] on board… [I've] felt frustration about that…—QI Team Physician 1 (Hospital B)
Staff spoke about the importance of empowering nurses as the ‘first line of defense’ to recognise opportunities for administering ANCS.…when they come in, our nurses know that [steroids] is on the radar for a preterm person, and if [physicians] don't say it quick enough, the nurse will prompt them.—QI Team 1 (Hospital F)
There are many times [nurses] ask, ‘Are we going to give steroids or not?’…This is their field of practice and they know very much when we administer them…—Provider 1 (Hospital C)
The role of teamwork in creating a culture and process that supports ANCS administration was also emphasised.It is a priority. Usually, it's a team approach. So if you have a really sick patient coming in, it's not just one nurse taking care of that patient. There's somebody grabbing and going and getting and shooting and documenting. I think it's just a priority.—QI Team 2 (Hospital D)
Administrative process emphasises reliability
OPQC sites had evidence of ANCS administration processes that include design features such as standardisation, reminders and redundancy. For example, staff spoke about the role of standard communication about ANCS, particularly during care transitions such as during nurse-to-nurse handoff or at the time of transfer from one facility to another.…we changed some of that [transport] paperwork… we recreated the med section to…gets the specific question ‘Were steroids administered already and what time?’…So that gets checked, so that kind of starts the trigger…—QI Team 3 (Hospital B)
I mean, just in the two years that I've been here, everyone has been really good about it, even someone that's been on antepartum for two months, every report of that person, and they had steroids on 7/20 and 7/21. It's communicated very, very clearly.—Nurse 1 (Hospital A)
Staff also spoke about the role of cues and prompts to heighten awareness about the importance of administering ANCS including order sets and posted reminders:…all the department order sets include corticosteroids. And I think that's important, because again, see an empty box, resident says; ‘I just want to check it.’—QI Team Physician 3 (Hospital A)
…kind of constantly in front of us…if you go into the board room [containing chalkboard listing the patients in active labor and their pertinent issues] and there's a preterm patient that is eligible, you'll see the [steroids] is on there—the date, the time for the repeat. It's there.—QI Team 4 (Hospital A)
In addition, at many sites, the process of administering steroids was designed to include redundancy in who identifies a woman as eligible for steroids because patients interact with a range of care providers and each of these encounters provides an opportunity to identify an eligible woman and administer ANCS.I think probably the common factor is that the perinatologist pretty much sees all those patients….if the OB hasn't ordered it, he's going to order it.—QI Team 1 (Hospital F)
I think we do a really good job because there's a lot of eyes watching. There's the nurses watching. There's the residents watching. There's the attendings watching. [QI Team Physician Lead] is watching. So everyone is looking to make sure …because we all know how important they [steroids] are.—QI Team 2 (Hospital D)
Administration process is timely and efficient
Another factor that was important to high rates of ANCS use was having a process that allowed for timely administration. Important aspects of the process were that the medication is readily available on the unit or is quickly obtained from the pharmacy:It's stocked here on the floor…It's in our Pyxis [Automated Medication Management System on the unit]. So we just ask a nurse for it even if it's an emergent situation…you assign one nurse the role of giving the steroids….—Provider 5 (Hospital D)
We call pharmacy and make sure they have them up here STAT. And we're on them constantly to get them to us as fast as possible since we are not able to store them in our Pyxis [Automated Medication Management System on the unit]. We can have them up here in 10 to 15 minutes.—Nurse 2 (Hospital B)
Additionally, sites had processes to ensure that providers are promptly available to assess patients in preterm labour and, subsequently, when the ANCS order is written, that the nurses receive verbal communication about the order:I guess that's nursing judgment for me…Doesn't matter me getting…her G's [Gravidity] and P's [Parity]. It's getting the doctor in there so we can get the steroids going.—Nurse 3 (Hospital C)
Multiple disciplines are involved
Engaging multiple disciplines in decisions about administration of ANCS and in discussions around interpreting the evidence emerged as an important theme. Enabling interactions between labour and delivery staff and the neonatal intensive care unit (NICU) team was emphasised:We never went to the NICU. We never got follow-up. So, we're really trying to change that…the more communication we get from the pediatricians, the more we feel like we're a team with them…the more effective we're going to be at making sure that we don't let some slip through the cracks.—QI Team Physician 2 (Hospital E)
So, again, that's another way of making sure things don't fall through the cracks. The neonatologist reviews the case and they write in their note that steroids were ordered and given.—Provider 2 (Hospital F)
Sites also emphasised engaging with the maternal-fetal medicine (MFM) specialists who typically care for high-risk pregnancies about interpreting and implementing the evidence for ANCS use and the eligibility of particular patients:We have [multiple] MFM's. And so, it's nice, to have like a sort of round table discussion about who's going to benefit and who's not. The pros, the cons, that kind of stuff.—QI Team Physician 1 (Hospital B)
Evidence of benefit supports use
In the context of interviews, nurses and physicians at participating hospitals restated the evidence for the important role of ANCS as part of their practice, suggesting the appreciation for the evidence-base is pervasive:I think steroids is one of the most important things because there are very few things in obstetrics that we can do to make pregnancies better. Steroids is one of the few things we have control of, our tocolytics don't work, we can't stop labour … Steroids help…there are some beautiful things in OB/GYN that tend to not be sexy anymore because we used it so long. Steroids is one of them.—QI Team Physician 3 (Hospital A)
I think it's extremely important. The literature shows that babies who go to special care spend less time in special care, spend less time on ventilators.—Nurse 6 (Hospital F)
Participants also emphasised the importance of educating others about the evidence behind ANCS administration, especially staff at referring community hospitals.If you point to the evidence, they'll [private practice physicians] follow the information, follow the trail.—QI Team Physician 1 (Hospital B)
…probably the biggest risk for not getting steroids in my opinion is patients being transferred from other institutions. They don't have the same kind of understanding of why steroids are important…But to really attack it and get to the referring physicians at other hospitals, we need to get out there even more than we have….—QI Team Physician 3 (Hospital A)
Having strategies to rapidly implement new guidelines as evidence changes is also important for achieving high rates of ANCS administration. For example, in early 2011, the American College of Obstetrics and Gynecology issued a new opinion supporting the safety of a single second rescue course of ANCS.24 Some providers spoke about how this new evidence aided in increasing rates of ANCS use because there was less concern about giving the first course if it could be followed by a second course should a woman remain undelivered:[I have] changed my opinion and practice based on what I've now been reading and what I've been taught…,we do know from the data we should not be giving [steroids] in numerous doses…but it's not harmful to have a rescue dose, and if that's indicated, I think you should give them. And, now based on at the end of my training, I have read and learned a lot more about that, and I feel differently now, I believe in rescue dose steroids.—Provider 1 (Hospital C)
When we started the project…we came down to [MFM physician]…about the rescue dose…And he was like, ‘We don't do that so much’…And we said, ‘Oh, well there's value to it.’’…we sat there and talked to him about it…we did talk about rescue dosing, was that something we were going to be doing.—QI Team 1 (Hospital F)
Benefit is recognised at all levels of the care team
Another key theme that emerged was the importance of recognising the benefits of ANCS for all levels of the care team including nurses, physicians, trainees and the patient themselves. Staff spoke about the importance of making ANCS a priority:It's a priority for everybody. That's one of the first things the doctors order, one of the first things that we do…—Nurse 5 (Hospital D)
It's expected of us as residents [trainees], and if we miss it, somebody is going to bring it up in the morning.—Provider 3 (Hospital B)
Physician leaders, particularly the obstetrical physician lead, were described as playing an important part in ensuring that the benefits of ANCS were recognised. Participants spoke of how the obstetrical champion emphasised the importance of ANCS among other physicians as well as nursing staff.…you have to have a physician champion to bring the providers along…and so [the obstetrical physician champion] is taking this information…out to the other practitioners that are here and saying, you know,…if you have a patient that you think is going to be preterm labour and perhaps deliver early, consider the steroids between this time and this time.—QI Team 6 (Hospital E)
…he's comfortable sitting and talking with nurses at the nurses’ station of why this person is getting steroids, why they wouldn't get steroids, we need to get it now, when they'll get it again…and the importance of it and the impact it could have on the baby. So he's good at the physician level and he's also good bringing it down, the information to the nursing staff.—QI Team 2 (Hospital D)
One particular hospital, with lower rates of ANCS (87%), spoke about their lack of a strong obstetrical leader and the impact this had on setting the tone regarding the importance of ANCS:I just wish we had a physician that was talking to the other physicians and saying, ‘Hey, you know, did you see that article? This is the new standard [single repeat ANCS course].’… That would be wonderful, and that's really in a perfect world, physician to physician.—QI Team 1 (Hospital F)
Although the care team is traditionally viewed as including only clinical personnel, participating sites also spoke about the important role of patients. With respect to ANCS, in a way, patients serve as extenders of the care team. By ensuring patients are aware of the benefits of ANCS, they can assume some responsibility for ensuring that, if eligible, they are appropriately treated:…patients are even educated on why they are getting it. They're not just like, ‘Oh, you're giving me a shot.’…I just had a patient last week that has had one course…and she was asking about her second course of steroids. She knows the importance of it and she knows that she was getting right at that 30-week mark again and wondering when she would be getting another course.—Nurse 1 (Hospital A)
I said ‘Did you know did you have steroids?’ And [patient] said, ‘No I didn't, but did ask my doctor about them.’—QI Team 1 (Hospital F)
Discussion
Our findings illustrate factors that have enabled reliable administration of ANCS at levels of >90% at OPQC hospitals. Comments provided by QI team members participating in the OPQC ANCS QI project, as well as insights from nurses and obstetrical providers not directly involved in these efforts, suggest that reliable implementation of ANCS is a function of: (1) presence of a high reliability culture, (2) processes that emphasise high reliability, (3) timely and efficient administration processes, (4) having multiple disciplines involved, (5) the strong foundation of evidence regarding the effectiveness of ANCS and (6) ensuring that benefit is recognised at all levels of the care team. Based on 2013 data from the Vermont Oxford Network registry of over 54 000 infants cared for in 950 NICUs, rates of ANCS administration among infants 24–33 weeks gestation averages 83% with the lowest quartile of centres having rates below 76%.25 Similarly, a recent study of over 33 000 infants in 120 hospitals in California found significant variation in ANCS rates with regional ANCS rates ranging from 68% to 93%.26 This suggests that there is still room for improvement and other centres could benefit from creating the processes and supporting conditions that enabled OPQC hospitals to achieve high rates of ANCS administration.
The major themes identified as supporting reliable ANCS administration at OPQC hospital are well aligned with known theories of implementation and high reliability. For example, the Promoting Action on Research Implementation in Health Services (PARiHS) framework suggests that for evidence to be implemented successfully there must be robust scientific evidence, a receptive context and appropriate facilitation of change.27–29 While the PARiHS framework is generally applied to the development of change strategies to increase evidence utilisation, our data suggest that, in the case of ANCS, the general elements of context, evidence and facilitation are also important in sustaining evidence delivery at high levels. Further, our data revealed nuances within each of these elements that appeared to be particularly important for sustaining evidence implementation as opposed to implementing a new standard. For example, with respect to evidence, our data suggest that the robustness of evidence is important, and the ability to modify and incorporate new evidence as it becomes available (eg, when new data surfaced about the safety of a single repeat course of ANCS) is also crucial.30 In addition, culture and physician leadership were cited as particularly important aspects of the local context with respect to sustaining ANCS administration at high rates. When looking to sustain a practice over long periods of time, these types of contextual factors that endure within an organisation or unit and are part of its general fabric appear to be important. Perhaps in cases of sustaining evidence implementation, it is less about a receptive context, and more about a culture that supports standardisation and high reliability.
Theories of reliability science and principles of high reliability organisations (HROs) that participants identified as supporting ANCS implementation at high rates included preoccupation with misses, empowerment of front-line staff and teamwork.10 ,11 The study of high reliability, defined as consistent performance at high levels over long periods of time, began with investigations of HROs in industries such as aviation and nuclear power.9 The key features of HROs identified by Weick and Sutcliffe, particularly preoccupation with failure, sensitivity to operations and deference to front-line expertise, are mirrored in the comments made by participants. Central to what was heard from study participants was the pervasive desire to eliminate misses, what is referred to in HROs as collective mindfulness.11 In addition, staff spoke about deference to front-line expertise when they discussed the central role of nurses in recognising opportunities to administer ANCS.
This study adds to the current understanding of conditions necessary for highly reliable implementation of evidence because our qualitative data point to important areas of overlap between these two separate bodies of theoretical knowledge—implementation and reliability science. Our data demonstrate where these theories from different disciplines can support and reinforce each other. For example, staff acknowledgement of the importance of administration processes that emphasise reliability combines components of facilitation, such as process improvement, standardisation and lean management, with principles from reliability science, including redundancy, reminders and standardised communication.10 ,31 Similarly, data suggesting the critical role of recognising the benefits of ANCS at all levels of the care team mix elements of the culture of collective mindfulness from HROs with features central to evidence-based practice, such as consensus across disciplines and opinion leaders/champions. Although improvement efforts focused on safety tend to draw on reliability principles and those focused on evidence-based guidelines tend to draw on implementation science theories, to date, we have seen little published work that draws on both disciplines despite calls for more transdisciplinary initiatives.32–34 For example, while implementation science and reliability science emphasise the importance of culture, the specific facets of culture that are emphasised differ between the two fields. Implementation science focuses more on creating a receptive culture through leadership and vision, data feedback, risk taking and emphasising learning.17 ,27 Reliability science emphasises preoccupation with failure, commitment to resilience and deference to front-line expertise.10 Organisations who are still struggling to implement evidence with high reliability should support an organisational culture that attends to all of these features. Similarly implementation researchers should draw on theories from both disciplines and think about these areas of overlap when designing interventions and creating supporting conditions to optimise evidence implementation. We believe that drawing on the synergies and components from these two disciplines can help create a solid scaffold for supporting the successful implementation of other types of evidence.
Among the sites that were examined in this study, there remained a small amount of variation in rates of ANCS use with Hospitals A and C having the highest rate of 97% and Hospital F having the lowest rate at 87%. Although our study was not designed as a quantitative or statistical analysis of the differences between higher and lower performing hospitals, some qualitative data suggested that sites with higher rates of ANCS seemed to have more formal methods of reviewing missed cases and stronger physician leadership. These observations should be studied further.
While this study contributes to our deeper understanding of the way evidence is implemented with high reliability, it has several limitations. First, OPQC centres were considered as having successfully implemented ANCS with high reliability because their rates of administration consistently exceeded 90%. However, some caution must be exercised. Rates much greater than 90% could represent overuse and reflect the practice of giving ANCS even when delivery is imminent, outside the window of maximal benefit, or in cases of false labour when a woman goes on to deliver at term. These are settings where steroids have not clearly been shown to be useful. Therefore, the factors we have identified that support high rates of ANCS use may also lead to some measure of overuse. Because traditional quality metrics for ANCS examine whether a woman receives at least one dose of ANCS prior to delivery,35 we used this measure for defining reliable implementation and exploring factors supporting implementation. Although OPQC hospitals initiated ANCS in >90% of eligible women, these women may have only received a partial course (one dose) or may have received ANCS outside of the ideal time frame of 2–7 days prior to delivery. Therefore, it is possible that additional factors are important in supporting ‘optimal’ ANCS administration whereby ANCS is administered in the best possible circumstances (eg, single course, administered 2–7 days prior to delivery). Lastly, our question guides were designed to elicit staff perceptions, thus some factors that staff identified as leading to high rates of ANCS administration may not have actually been the true contributors.
Although achieving rates of ANCS use among OPQC hospitals of 93% (average) still means that there is nearly a 1 in 10 chance that suboptimal care will be provided. The achievement of >90% reliability indicates that specific common processes and supportive conditions exist for the application of evidence in practice at participating sites with some room remaining for site-based interpretation of the guidelines to minimise overuse.10 The details of the process and conditions supporting ANCS administration revealed through this qualitative study are supported by reliability and implementation science and have been assembled in a toolkit to support practical application by centres who still have opportunities to improve ANCS administration.36 In addition, these six major themes are also likely to be important for implementation of other evidence-based practices at high levels of reliability. Future research should test the effectiveness of interventions that incorporate these components, informed by reliability and implementation science, in increasing the implementation of evidence in practice.
Acknowledgments
We would like to thank the nurses, physicians, and hospital staff who took time out of their busy schedules to speak with us and share their exciting work. We would like to thank the Ohio Perinatal Quality Collaborative leadership and staff for their ongoing feedback and support, particularly Sandra Fuller, Katherine Clarke-Myers, Jay Iams, and Barbara Rose. We would also like to thank Joshua Watring and Abigail Chandler for their assistance in the logistics of this project and Pamela Schoettker for her editorial assistance with this manuscript.
References
Footnotes
Contributors HCK led the design, data analysis, interpretation and drafting of the manuscript. SNS, LG, and HCK developed the interview guide and conducted all interviews. HCK, SNS and CC conducted qualitative data analysis. All authors participated in the conception and design of the study, interpretation of data and critical revisions of the manuscript. All authors have granted final approval of the version to be published.
Funding The OPQC ANCS Project was funded by the Ohio Department of Health, the Ohio Department of Medicaid, and the Centers for Disease Control (State-Based Perinatal Quality Collaborative programme, CDC-RFA-DP11-111003CONT13). Support for this research was provided by cooperative agreement U19HS021114 from the Agency for Healthcare Research and Quality. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality.
Competing interests None declared.
Ethics approval Cincinnati Children's Hospital Medical Center IRB determined this study to be exempt.
Provenance and peer review Not commissioned; externally peer reviewed.