Factors driving differences in the adoption of quality management practices among hospitals: A two-phase, sequential mixed-methods analysis

Background Although all hospitals aim to deliver high-quality care, there is considerable variation in their adoption of quality management (QM) practices. Organizational and environmental factors are known to drive strategic decision-making in hospitals, but their impact on the adoption of QM practices remains unclear. Purpose Our study aims to identify multiple organizational and environmental factors that explain variation in the adoption of QM practices among hospitals and to explore mechanisms underlying these relationships. Methodology We conducted a two-phase, sequential mixed-methods study of German acute care hospitals. The quantitative phase used between-effects regressions to identify factors explaining variation in the number of QM practices adopted by hospitals from 2015 to 2019. The qualitative phase used semistructured interviews with quality managers to gain in-depth insights. Results The number of QM practices adopted by a hospital was significantly associated with factors like hospital size and the presence of an emergency department or QM steering committee. Our qualitative findings highlighted potential mechanisms such as the presence of an emergency department serving as a proxy for organizational complexity or urgency of case-mix. Conclusion We provide an overview of factors driving QM adoption in hospitals, extending beyond the focus on single factors in previous research. Future studies could explore additional factors highlighted by our interviewees. Practice Implications Our results can inform interventions to strengthen QM in hospitals and guide future research on this topic.


I
n many high-and middle-income countries, hospitals face increasing competition and growing demands for quality transparency, putting them under pressure not only to provide but also to demonstrate high-quality patient care (Botje et al., 2014;Hammer et al., 2013).Along with the public discourse on quality and patient safety, as well as increasingly demanding regulatory requirements, this has contributed to the widespread adoption of quality management (QM) practices in hospitals to improve the quality of processes and outcomes (Hammer et al., 2013;Kristensen et al., 2015).
QM can be defined as "a set of interacting activities, methods, and procedures used to monitor, control, and improve the quality of care" (Hammer et al., 2013, p. 2) and thus encompasses a wide range of practices in areas such as clinical risk management, critical incident reporting, and infection control (e.g., Botje et al., 2014;Hammer et al., 2013;Zarei et al., 2019).Despite the widespread adoption of QM in hospitals, the extent and quality of its implementation vary widely (Hammer et al., 2013;Seyfried et al., 2022;Wagner et al., 2006).From previous literature, we know that organizational and environmental factors drive strategic decisionmaking in hospitals (e.g., Jennings et al., 2019;Kazley & Ozcan, 2007;Schneider et al., 2021).However, the extent to which this observation applies to the strategic decision to implement QM practices remains unclear, with no studies, to our knowledge, having addressed this topic in a comprehensive way to date.Although prior hospital management research on QM has consistently identified factors that influence the adoption of QM practices, it has tended to focus on single factors only, leaving the information scattered throughout the literature and making it difficult to disentangle single effects while accounting for other variables.For instance, it is unclear whether differences in QM adoption across hospital ownership types might be driven by differences in hospital size, or whether it is the degree of urbanization or the degree of competition that drives variation.As a result, it remains difficult for hospital management and policymakers to identify the factors that have the potential to influence the use of QM in hospitals.
To address this gap in the literature, the present study employed a two-phase sequential mixed-methods design to answer the following research question: Which organizational and environmental factors drive variation in the adoption of QM practices in hospitals, and what are some of the underlying mechanisms?Although this type of research question has traditionally been investigated only through quantitative methods, we chose a mixed-methods approach based on the rationale that this can lead to a better understanding of complex phenomena such as QM in hospitals (Fetters et al., 2013).In the quantitative phase, our objective was first to derive a set of organizational and environmental factors based on prior evidence and expert discussions and to develop hypotheses regarding their potential effects on the adoption of QM practices.We then tested our hypotheses using between-effects regression and panel data from a large sample of German acute care hospitals from 2015 to 2019.Subsequently, in the qualitative phase, we conducted a range of semistructured interviews with quality managers from a heterogeneous subset of the hospitals in the sample used for our quantitative analysis.Our objective in this phase was to illustrate, clarify, and elaborate upon the results from the quantitative phase in order to gain a deeper understanding of the underlying mechanisms and path dependencies of the identified factors.This involved exploring the expectations of practitioners, identifying explanations for the effect of the identified factors on the adoption of QM practices, and highlighting any additional or unexplored factors.
We found that the number of QM practices used by a hospital was significantly associated with factors like hospital size, the presence of an emergency department or QM steering committee, and hospital location.Our qualitative findings shed light on potential mechanisms underlying these relationships and offer explanations for the findings that were not in line with our theoretical expectations.The qualitative findings also point to additional and unexplored factors, such as top management's motivation for and support of QM, the quality manager's scope of authority and standing in the hospital, and the degree of cooperation within and among hospitals.With these results, we make two main important contributions to hospital QM theory and the literature.First, we extend the scope of previous work by providing a more comprehensive analysis of differences in hospitals' adoption of QM practices, both in terms of the number of factors investigated and the underlying mechanisms through which they might exert an effect.Second, we add to the literature by identifying additional and unexplored factors, revealing valuable information for both research and practice.Overall, our findings point to ways in which incentive systems could be designed or refined to enhance hospitals' use of QM.

Theory and Hypothesis Development
In general, differences in the strategic behavior of organizations can be explained by environmental and organizational factors, as well as by factors at the level of individual decision-makers.Theoretical frameworks such as institutional theory (DiMaggio & Powell, 1983;Meyer & Rowan, 1977), resource dependence theory (Pfeffer & Salancik, 1978), contingency theory (e.g.Lawrence & Lorsch, 1967;Thompson, 1967), and strategic choice theory (Child, 1972) offer different perspectives on how these factors drive variation in strategic behavior and how these factors might interact.Although institutional theory focuses on the external pressures that are exerted on organizations and to which they must respond in order to gain legitimacy (DiMaggio & Powell, 1983), resource dependence theory assumes that organizations depend on the resources within their environments.To secure the flow of resources, organizations may, for example, adapt their structures or behavior to meet changing environmental conditions or the demands of resource providers (Pfeffer & Salancik, 1978).Through the lens of contingency theory, organizations are exposed to a variety of internal and external environmental influences (Shepard & Hougland, 1978) and must respond to changes in their environment to restore fit between a hospital and its context (Donaldson, 2001).With the additional consideration of decision makers' strategic choices (Child, 1972), the principles of strategic choice theory extend contingency theory by providing a more comprehensive explanation of organizations' responses to changes in their internal and external environments.
With these theoretical frameworks in mind, we posited that the organizational and environmental factors that differ between hospitals can explain systematic variation in their adoption of QM practices.We therefore developed a preliminary set of factors to include as independent variables in the quantitative phase of our study based on a review of the related literature (Alexander et al., 1993;Alharbi & Yusoff, 2012;Ballantine et al., 2008;Botje et al., 2014;Eldenburg et al., 2004;Goldstein & Naor, 2005;Hammer et al., 2013;Jennings et al., 2019;Kazley & Ozcan, 2007;Leggat & Balding, 2019;Patidar et al., 2017;Potter & Dowd, 2003;Rotar et al., 2016;Schneider et al., 2021;Wagner et al., 2001Wagner et al., , 2014;;Zarei et al., 2019).Subsequently, we discussed the factors in a series of consensus meetings among all authors and with several experts in the research field, yielding a final list of independent variables to be used in our regression analyses.All of the authors and the participating experts had experience in conducting and evaluating studies that investigate variation in the strategic decisions of hospitals, such as decisions involving QM certification, human resource management practices, privatization and corporatization decisions, and strategy choices.The organizational variables on the final list were hospital size, teaching status, ownership type, level of specialization, nurse staffing levels, presence of an emergency department, the inclusion of the quality manager on the management board or within the medical directorate, the presence of a QM steering committee, and the gender of the quality manager.The environmental variables were the degree of hospital competition in a district, hospital location (i.e., Rationale: Larger hospitals may have better access to internal and slack resources, such as additional staffing, technology, financial resources, and diverse interorganizational relationships, which are crucial for investing in the implementation of strategic decisions like QM.

2
Teaching hospitals Hypothesis: Teaching hospitals will use more QM practices than nonteaching hospitals.
Rationale: Teaching hospitals are more likely to be proactive, invest in quality goals and patient safety, and have a culture of innovation in clinical practice, increasing the use of QM.

3
Ownership types Hypothesis: Private for-profit hospitals will use more QM practices than hospitals of other ownership types.
Rationale: Compared to private non-profit and public hospitals, for-profit hospitals probably have more financial resources to invest in QM practices, are more likely to view these practices as a means to increase efficiency and attract a steady influx of patients, and have a stronger focus on quality and more professionalized structures, facilitating the implementation of QM practices.
4 Specialization Hypothesis: Hospitals with a higher degree of specialization will use more QM practices than hospitals with a lower degree of specialization.
Rationale: A higher degree of specialization in hospitals leads to increased medical and administrative attention towards service provision, which translates into an increased emphasis on quality management, including the use of a larger number of QM practices.

5
Nurse staffing Hypothesis: Hospitals with higher nurse staffing levels will use more QM practices than hospitals with lower nurse staffing levels.
Nurses play a crucial role in implementing QM.A higher level of nurse staffing allows QM practices to be implemented more comprehensively.

Presence of emergency department
Hypothesis: Hospitals with emergency departments will use more QM practices than hospitals without emergency departments.
Rationale: Hospitals with emergency departments have a higher share of emergency cases, for which standardized and documented QM practices are especially important.Moreover, these hospitals tend to be more complex, increasing the need for QM practices as a means to monitor and maintain quality through standardization.
7 Quality manager's gender Hypothesis: Hospitals whose quality manager is female will use more QM practices than hospitals whose QM manager is not female.
Rationale: Women tend to show more altruism, risk aversion, and aversion to inequality, which could lead to female quality managers promoting a more comprehensive approach to QM.This could be driven by a desire to minimize any risks related to quality issues and to maximize the quality of care as an expression of altruism.
8 Quality manager is a member of the medical directorate or management board Hypothesis: Hospitals whose quality manager holds a position in the medical directorate or management board will use more QM practices than hospitals whose quality manager does not have this dual affiliation.
Rationale: When the quality manager holds a position in the medical directorate or the management board, he or she is likely to have greater influence in prioritizing quality issues in the top management's agenda.9 Presence of a QM steering committee Hypothesis: Hospitals with a QM steering committee will use more QM practices than hospitals without a QM steering committee.
Rationale: The formal integration of QM through structures like a dedicated QM department or steering committee is likely to strengthen the standing and the power of QM in an organization.Such a committee can assume key roles, such as developing strategic QM plans, providing quality data and information, and developing quality standards.
Note.Because of space limitations, only the main rationale is specified for each factor.Detailed hypotheses and rationales, including references to theories and prior literature, are presented in Supplemental Digital Content 7, http://links.lww.com/HCMR/A147.
the state), degree of urbanization, and socioeconomic characteristics of the region (i.e., environmental munificence).Table 1 presents our hypotheses for the organizational factors, and Table 2 presents our hypotheses for the environmental factors.

Study Design
We employed a two-phase, sequential quantitative to qualitative mixed-methods design.First, to identify multiple organizational and environmental factors that drive variation in the adoption of QM among hospitals, we modeled the effect of our list of predetermined factors on the total number of QM practices used by each hospital in our sample by employing between-effects regressions.The objective of the quantitative phase was to test whether these factors systematically explained variation in the adoption of QM practices in the expected direction.Second, to illustrate, clarify, and elaborate upon the results of the quantitative phase, we conducted qualitative, semistructured interviews with quality managers from a heterogeneous subset of these hospitals.By using the qualitative tradition of case studies (Creswell, 1998), we sought to explore the expectations of practitioners in order to obtain in-depth insights into the identified associations and to highlight any additional or unexplored factors, thus extending the results of the quantitative phase.
Our mixed-methods approach involved integrating our quantitative and qualitative methods through two main processes: connecting and building.The connecting process entailed using the sample of hospitals from our quantitative data analyses to select the qualitative sample.The building process involved developing the interview guide for the expert interviews based on the findings of the quantitative data analyses (Fetters et al., 2013).

Quantitative Phase
Setting and data.We constructed a set of panel data on German acute care hospitals from 2015 to 2019 by drawing upon two sources: (a) the structured quality reports that German hospitals must publish annually and which are based on structural and performance data collected by a governmentadministered survey and (b) the INKAR database of the Federal Institute for Research on Building, Urban Affairs and Spatial Development (BBSR) to obtain information on regional characteristics.The data were subjected to extensive validity and plausibility checks.To ensure comparability among hospitals in our sample, we excluded university hospitals, psychiatric hospitals, day hospitals, rehabilitation centers, and hospitals with fewer than 30 beds from our analyses.Our final sample comprised 7,454 hospital-year observations from 1,571 individual hospitals.Differences in state health policies, structures for health care provision, population demographics and preferences, as well as varying funding rules and capacities across the German states, are likely to lead to diverse emphases, resources, and needs for QM practices in hospitals.
12 Degree of urbanization Hypothesis: Hospitals in more urbanized areas will use more QM practices than hospitals in less urbanized areas.
The degree of urbanization might influence QM implementation insofar as QR practices potentially diffuse more quickly in urbanized areas, where the density of organizations is greater than in rural areas, and by affecting the amount of resources available for QM, which are also higher in more urbanized areas.

Environmental munificence
Hypothesis: Hospitals located in more munificent districts will use more QM practices than hospitals located in less munificent districts.
Organizations located in more munificent environments have the advantage of easier access to the necessary resources and the ability to build interorganizational relationships, facilitating the implementation of management decisions such as those to use QM.
Note.Because of space limitations, only the main argument is specified for each factor.Detailed hypotheses, including references to theories and prior literature, are presented in Supplemental Digital Content 8, http://links.lww.com/HCMR/A148.
Definition and measures.We operationalized the adoption of QM practices in a hospital as the total number of QM practices a hospital used during a year.We used the structured quality reports to calculate this number and form an additive QM index.This served as the dependent variable in our study and ranged from 0 to 25.We only counted QM practices that were available for all years of our study period and that were based on recognized expert guidelines and recommendations for action from relevant national or international professional societies.Overall, our index included practices from the QM areas of clinical risk management (17 practices), infection control (five practices), and critical incident reporting (three practices).All QM practices and the share of hospitals indicating they employed this practice per year are presented in Supplemental Digital Content 1, http://links.lww.com/HCMR/A141.Because the adoption of each QM practice was represented by a dichotomous variable, we tested the reliability of the QM index by means of the internal consistency method using the Kuder-Richardson formula 20 coefficient (KR-20).The KR-20 for the QM index was .92,indicating high internal consistency (Ferketich, 1990).Data on the organizational and environmental factors hypothesized to explain variation in QM practices were taken from, or calculated based on, the structured quality reports.Hospital size was measured in units of 100 beds.Academic teaching status was a dichotomous variable indicating whether a hospital had this status.Ownership type was measured by the three categories: "public," "private for-profit," and "private non-profit."Hospital specialization was measured using the Gini coefficient, as proposed by Daidone and D'Amico (2009).The Gini coefficient was scaled to an interval ranging from 0 to 100, with a higher Gini coefficient indicating a more specialized hospital.Nurse staffing was measured by the nurseto-bed ratio, which was calculated as the number of full-time equivalent nurses divided by the number of acute care beds.The presence or absence of a 24-hour emergency department was measured by a dichotomous variable.The quality manager's gender was recorded as "male," "female," or "no information," with "female" being the reference category."No information" applied to quality managers who did not report their gender in the structured quality reports.Two dichotomous variables indicated whether the quality manager was a member of the medical directorate or management board.Lastly, a dichotomous variable indicated whether hospitals had a QM steering committee in place.
Regarding our four environmental factors, we used the Herfindahl-Hirschman index as a proxy for hospital competition in the district in which the hospital was located.It was calculated based on inpatient discharges and is presented as a percentage, thus taking values between 0 and 100, with higher values indicating a less competitive market.Hospital location was represented by the German state in which the hospital was located.Data on the degree of urbanization of the hospitals' location at the district level were added from the database of the BBSR and were classified as "large city," "urban district," "rural district," or "remote district."To capture the regional socioeconomic environment (i.e., the environmental munificence), we used the district-level German Index of Socioeconomic Deprivation (GISD) developed by Kroll et al. (2017).The GISD score uses eight indicators from the database of the BBSR that relate to the three core dimensions of social inequality: education, occupation, and income.A higher GISD score indicates a higher level of socioeconomic deprivation (Kroll et al., 2017).Associations between the study variables are presented in Supplemental Digital Content 2 (see Table, Supplemental Digital Content 2, http://links.lww.com/HCMR/A142).
Data analysis.All statistical analyses were carried out using Stata/SE17 (StataCorp, 2021).In panel data, variation in the number of QM practices can be decomposed into two parts: that between hospitals and that within hospitals over time.Because the objective of the quantitative phase was to examine between-hospital variation in QM, we exploited the between-variation in our data by estimating a betweeneffects regression model.A between-effects model relies solely on the cross-sectional information in the data by computing the hospital-specific means across the study period for each variable.Bootstrapped standard errors with 500 repetitions were used to control for heteroskedasticity.To account for any within-variation, we ran a random-effects within-between model using the xthybrid command (Schunck & Perales, 2017) in Stata as an additional sensitivity analysis.In addition, we calculated three separate indices for the three dimensions of QM practices-clinical risk management, infection control, and critical incident reporting-and used these as dependent variables.Furthermore, we checked the consistency of our results by using the individual QM practices as dependent variables.

Qualitative Phase
Two of the authors were involved in qualitative data collection and analysis.The first had extensive expertise in developing and implementing sampling strategies, as well as in interview guideline development, data collection and interviewing, qualitative data analyses and coding techniques, and validity measures.The second, who conducted the interviews, had taken part in a range of courses on designing and conducting interviews and on coding qualitative data.
Sampling and data collection.The qualitative phase consisted of semistructured online and telephone interviews with quality managers from a heterogeneous subset of the hospitals in the sample used in our quantitative analysis.To identify interviewees, we followed a voluntary response sampling approach.We sent invitations by e-mail to quality managers of all hospitals in the quantitative data sample for whom e-mail addresses were available (N approx = 800), with a request to forward the e-mail to other potentially interested and suitable colleagues.In total, 41 quality managers accepted the invitation to participate and provided informed consent.The interviews were conducted by one of the authors between April and May 2022.Because some hospitals or hospital systems employed several individuals involved in QM who expressed a desire to collaboratively share their experiences, six of these interviews took place with two interviewees each and one interview with three interviewees, resulting in 41 interviewees who took part in a total of 33 interviews.We confirmed heterogeneity in the sample by assessing the characteristics of the hospitals of the responding quality managers, such as hospital size, ownership type, and total number of QM practices.Because information saturation was achieved after 33 interviews, we refrained from sending additional reminders.
Upon accepting the invitation to participate, interviewees were provided with a study information sheet.Additionally, 2 days before each interview, interviewees received a questionnaire covering general information (e.g., sociodemographic factors), as well as a summary of the results of the quantitative phase of the study.In the interviews, we presented the interviewees with each result and asked them whether it was something they would or would not have expected and why.The aim of doing so was to obtain deeper insights into the results.For example, in cases where a quantitative finding was in line with our hypotheses, this entailed assessing whether the explanations provided by the interviewees matched our theoretical arguments.Conversely, in cases where a quantitative finding was contrary to our hypotheses, interviews provided insight into why our hypotheses were not supported.We also aimed to identify additional and unexplored factors in the interviews.Before we started collecting data, the study protocol received ethics approval from the responsible board of the investigating organization.The interviews were recorded and transcribed using Amberscript (Amberscript Global B.V., 2022) and then corrected manually.
Data analysis.Two researchers analyzed the data in a collaborative coding process to ensure validity (Tracy, 2010).The first researcher was involved in all aspects of preparing the fieldwork and collecting data.The other researcher held an outsider position and scrutinized the analysis and its results.Both researchers independently worked with the data following the main steps of analysis described below.After obtaining preliminary results, we engaged in detailed discussions on each part of the analysis.This enabled us to compare different perspectives on the data and discuss possible explanations for the phenomena we had identified, ultimately leading to a joint interpretation of the data.
We analyzed data only after data collection had ended, coding the transcribed interviews using MAXQDA 2022 (VERBI Software, 2021) based on the two-step coding process (basic coding and fine coding) proposed by Rädiker and Kuckartz (2020).In the basic coding step, we assigned codes corresponding to the investigated environmental and organizational factors (one code for each factor, e.g., "competition") and one code for "additional factors."After completing the basic coding step, we conducted the fine coding step, during which we assigned the subcodes "interviewees' expectations" or "explanation of quantitative findings" to each unit of meaning coded as an environmental and organizational factor.The subcode "interviewees' expectations" was assigned to any unit of meaning in which the interviewees indicated whether they expected a factor to be related to hospitals' QM practices and, if so, in which direction.The subcode "explanation of quantitative findings" was assigned to units of meaning that covered any explanation for the relationship or lack of relationship between a factor and the number of QM practices used by hospitals.Finally, for each unit of meaning coded as an additional factor that interviewees thought might contribute to differences in QM between hospitals, we inductively generated subcodes such as "financial resources," "top management support," and "interprofessional collaboration" (not examined in our quantitative analyses).

Quantitative and Qualitative Samples
The characteristics of the quantitative sample in terms of distributions, means, and standard deviations (overall as well as decomposed into between and within variation) of the study variables are presented in Supplemental Digital Content 3 (see Table, Supplemental Digital Content 3, http://links.lww.com/HCMR/A143).To examine the representativeness of our quantitative sample, we compared it to the German hospital market by size and ownership.With an average of 267 beds, the hospitals in our quantitative sample were roughly representative in terms of size (compared to 258 beds in Germany).As for ownership, private for-profit hospitals were underrepresented (24%, compared to 38% in Germany), whereas public hospitals were overrepresented (34%, compared to 29% in Germany).Private non-profit hospitals were also overrepresented (42% compared to 34% in Germany).Overall, our quantitative sample largely reflects the German hospital market.Although there were deviations that need to be considered, they are of minor importance for our research question.
As noted above, our qualitative sample consisted of 33 interviews with 41 quality managers.The characteristics of the interviewees are presented in Supplementary Digital Content 4 (see Table, Supplementary Digital Content 4, http://links.lww.com/HCMR/A144).The interviewees came from hospitals with a bed capacity of less than 100 to more than 1,000 beds and of all ownership types, with public and private non-profit hospitals being equally represented and private for-profit hospitals less common in the sample (five out of 33).The QM index in our sample ranged from 18 to 25, with most QM indices above 20.As a result, our qualitative sample includes some heterogeneity, allowing us to provide insights into hospitals with varying backgrounds.

Quantitative and Qualitative Results
The results of the quantitative phase are given in Table 3. Overall, the organizational and environmental factors examined in our study jointly explained 25% of variation between hospitals in the number of QM practices used, as shown by the between-value of the R 2 (see Table 3).In the sensitivity analysis in which we estimated a random-effects withinbetween model, the results remained stable (see Table, Supplemental Digital Content 5, http://links.lww.com/HCMR/A145).In the sensitivity analyses using the three QM dimensions as dependent variables, the results also remained largely stable, with only a few changes in significance in one of the three dimensions (e.g., specialization, QM steering committee, and quality manager's gender being nonsignificant in one of the three regressions, and private for-profit ownership being significantly positive in one regression; see Table, Supplemental Digital Content 6, http://links.lww.com/HCMR/A146).At the level of the distinct QM practices, the results remained supportive of our overall findings but were more heterogeneous.
In the following sections, we apply the weaving approach for mixed-methods designs proposed by Fetters et al. (2013) and report the results from the quantitative and qualitative phases together, factor by factor.In Tables 4 and 5, the summarized explanations of the results are illustrated using selected quotes from the interviews.

Organizational Factors
Size.The results of our quantitative analysis show that hospital size was significantly positively associated with the number of QM practices adopted by a hospital.In other words, larger hospitals used more QM practices than smaller ones (see Table 3), supporting Hypothesis 1.In our interviews, several interviewees felt that this finding was to be expected and explained this sentiment by noting, for example, that larger hospitals tend to be more complex and therefore need a For the nominal variables "ownership type" and "manager's gender," the interviewees considered the hypothesized direction of the association (as specified in square brackets in the first column).Quotes were selected from the interviews based on their relevance to the research objectives (the extent to which they provided meaningful insights into the findings of the quantitative phase), whether they were representative of the broader data set (capturing the diversity of perspectives or experiences in the study population), clarity and precision (conveying the intended message), and impactfulness (uniqueness or depth of insight provided).
to use more QM practices to structure the increased number of processes (e.g., Interview 16; see Table 4).In addition, interviewees noted that larger hospitals may (a) have more financial, human, and intellectual resources; (b) place a higher value on QM; or (c) face more external requirements and therefore have a greater need for QM practices (e.g., Interviews 5 and 19).However, several interviewees said they would not have expected there to be differences and remarked that, due to regulatory requirements, QM ought to be the same in every hospital, regardless of its size (e.g., Interviews 17 and 20).
Academic teaching status.According to our quantitative findings, teaching hospitals used significantly more QM practices than did nonteaching hospitals (see Table 3), supporting Hypothesis 2. Several interviewees indicated that this finding was to be expected and elaborated that teaching hospitals have (a) a greater need for QM practices to structure and standardize processes for teaching purposes, (b) greater awareness of and interest in QM due to teaching activities, (c) a different structure compared to nonteaching hospitals (more complex, larger), and (d) more external requirements such as from their affiliated universities and therefore have a greater need for QM practices (e.g., Interview 32).In contrast, the interviewees who said they would not have expected there to be differences explained their sentiment by stating that the quality requirements for teaching hospitals are the same as for other hospitals (Interview 19; see Table 4).Note.Quotes were selected from the interviews based on their relevance to the research objectives (the extent to which they provided meaningful insights into the findings of the quantitative phase), whether they were representative of the broader data set (capturing the diversity of perspectives or experiences in the study population), clarity and precision (conveying the intended message), and impactfulness (uniqueness or depth of insight provided).
Ownership.In the quantitative analyses, private non-profit and private for-profit hospitals did not differ significantly from public hospitals in the number of QM practices they used.Thus, Hypothesis 3 is not supported by the quantitative data.In the interviews, several interviewees indicated that this quantitative finding is not what they would have expected and that they would have assumed that private forprofit hospitals would be more structured due to their profit orientation and therefore would have had more QM practices in place (e.g., Interviews 10 and 26).In contrast, another group of interviewees indicated that the quantitative finding is indeed what they would have expected and attributed this to (a) hospitals of different ownership types having the same requirements, (b) ownership not playing a role in everyday hospital operations, and (c) other factors having more influence (e.g., Interview 32; see Table 4).
Specialization.Our quantitative findings showed that more specialized hospitals used significantly fewer QM practices than did less specialized hospitals (see Table 3).Although this finding does not support the direction of effect postulated in Hypothesis 4, the magnitude of the effect was small.Most interviewees said they would have expected specialization to be positively related to QM practices, citing, for example, the fact that specialization is often based on certification, which is commonly associated with a strong QM system (e.g., Interview 2).Nevertheless, one explanation for our quantitative finding according to the interviewees could be that more specialized hospitals have less need for QM practices due to the potentially smaller range of services they offer, which in turn is associated with higher standardization, lower complexity, and lower number of interfaces with other departments (e.g., Interview 9).
Nurse staffing.Our quantitative results showed that hospitals with higher nurse-to-bed ratios used significantly more QM practices compared to hospitals with lower nurse-to-bed ratios (see Table 3), supporting Hypothesis 5.The majority of interviewees indicated that this finding was to be expected, mentioning that nurses in hospitals with higher nurse-to-bed ratios have more time for QM, whereas nurses in hospitals with lower staffing levels are stretched thin with their daily workloads (e.g., Interviews 6 and 23).In addition, they indicated that hospitals with higher levels of nurse staffing have a greater need for monitoring and coordination (e.g., Interview 6).
Presence of an emergency department.Our quantitative findings showed that hospitals with a 24-hour emergency department used significantly more QM practices than those without (see Table 3), supporting Hypothesis 6.Interestingly, only a few interviewees said this finding is what they would have expected, mentioning that emergency departments are critical departments with many interfaces with other departments and high patient turnover, which may increase the need for QM practices in order to standardize processes (e.g., Interview 32).
In addition, interviewees noted that hospitals with emergency departments are subject to greater external pressure because policymakers and medical societies impose more standards on them (e.g., Interview 9).However, most interviewees told us that the empirical finding is not what they would have expected, noting, for example, that there are few QM practices specific to emergency departments (e.g., Interview 4).
Quality manager's gender.Our quantitative findings showed that hospitals with a female quality manager used significantly more QM practices than hospitals with a male quality manager (see Table 3), supporting Hypothesis 7.Only a few interviewees said they would have expected this finding, with one participant suggesting that women tend to be less assertive than their male counterparts and therefore need QM practices to justify their work to hospital management.In addition, it was also mentioned by male and female interviewees alike that women tend to be more structured and organized and thus to use QM practices as a tool.Most interviewees, however, indicated that our quantitative finding is not what they would have expected, as they would not have thought that gender differences would have any relationship with the adoption of QM practices.
Quality manager being a member of the medical directorate or management board.Our quantitative findings showed that hospitals in which the quality manager was a member of the management board or medical directorate did not differ significantly in terms of the number of QM practices they used versus hospitals in which quality managers did not have this dual affiliation.Thus, Hypothesis 8 is not supported.Several interviewees told us they would have expected a positive association, arguing that if the quality manager is part of the management board or medical directorate, this means that there is an awareness of the fundamental importance of QM at the highest levels of management and thus resources are more likely to be made available (e.g., Interview 27).However, most interviewees said they would indeed have expected this factor to be unrelated to the adoption of QM practices, noting that (a) the quality manager alone is not able to achieve fundamental changes and (b) members of the medical directorate and the management board all have a similar level of competence, that is, it is the position of the quality manager and not the dual role that is decisive for the adoption of QM practices (e.g., Interview 31).
QM steering committee.Our quantitative findings showed that hospitals that had a QM steering committee used significantly more QM practices than hospitals without a QM steering committee (see Table 3), supporting Hypothesis 9. Most interviewees told us that this finding aligned with what they would have expected, arguing that the presence of a QM steering committee can be considered an indicator of the priority given to QM in a hospital (e.g., Interview 4).Interviewees also suggested that hospitals with a QM steering committee take a more systematic approach to adopting and implementing QM practices.They additionally pointed out that a QM steering committee also acts as a multiplier, involving a larger group of people in QM, thus ensuring stronger obligations and more pressure for action (e.g., Interview 29).

Environmental Factors
Competition.According to our quantitative findings, competition was not significantly associated with the number of QM practices used in hospitals.Thus, no support for Hypothesis 10 could be found.The majority of interviewees said they would have expected a relationship between competition and QM practices, arguing, for example, that hospital certification, which in Germany is based mostly on evaluations of a hospital's QM system, plays a role in competition.That is, if some hospitals in the market have certain certifications, other hospitals may feel the need to seek certification as well, which in turn could affect the number and type of QM practices used (e.g., Interview 18).However, a few interviewees said it did not surprise them that QM remained unaffected by competition, attributing this finding to (a) QM not being perceived as a competitive advantage from the patients' or hospital management's point of view, (b) competition being difficult to measure (considering that it is comparatively low in some areas), or (c) QM already being well established, meaning that it cannot be changed easily (e.g., Interview 16).
Location.Our quantitative findings showed that hospitals differed significantly in the number of QM practices they used depending on the state in which they were located (see Table 3), supporting Hypothesis 11.Only a few interviewees said they would have expected differences across states, offering explanations such as (a) differences in the competitive structures of the states, (b) Germany's historical development, (c) different levels of hospital investment, or (d) different types of hospitals in the respective states (e.g., Interviews 3, 9, and 12).
Degree of urbanization.According to our quantitative findings, hospitals in rural areas used significantly fewer QM practices than those in large cities (see Table 3), supporting Hypothesis 12.
Only very few interviewees said they would have expected the degree of urbanization to be related to QM practices.Regardless of their point of view, most interviewees offered explanations for our quantitative finding, noting that urban hospitals might (a) have a greater willingness for development, for example, in terms of QM; (b) have greater financial resources and expertise; (c) be under greater competitive pressure; or (d) have a different hospital structure compared to hospitals in rural areas, and therefore have a greater need for QM practices, for example, to cope with increased complexity and scale (e.g., Interview 26).

Socioeconomic condition/environmental munificence.
Our quantitative findings indicate that hospitals in areas with higher socioeconomic deprivation did not differ significantly from hospitals in areas with lower socioeconomic deprivation with regard to the number of QM practices they used.Thus, Hypothesis 13 is not supported.Most interviewees said that this finding is what they would have expected.They attributed the lack of significance to (a) the provision and quality of health care being independent of the population served and (b) the socioeconomic conditions of a region not reflecting the motivation and qualifications of hospital staff, that is, the people who apply or promote QM practices (e.g., Interview 19).

Additional Organizational and Environmental Factors
Interviewees identified a number of factors they believed could contribute to differences in QM beyond those examined in the quantitative phase of our study.The majority of organizational factors mentioned were directly related to QM itself, such as the financial and human resources of the QM department, length and continuity of QM efforts, composition and meeting frequency of the QM steering committee, how the QM was embedded in the hospital, qualifications and motivation of the quality manager, and the quality manager's scope of authority and standing in the hospital.For instance, it was noted that the availability of a separate QM budget is a decisive factor: Other organizational factors were general hospital characteristics, such as human resources and the financial resources of the hospital, the hospital's case mix, its number of certifications, membership in a hospital system, as well as factors related to the hospital's internal processes, such as interprofessional collaboration within the hospital.
A factor I feel was overlooked [in your analysis] but think is very important is the hospital's annual performance.[…] My hypothesis would be: the poorer the financial results, the less emphasis on QM.For environmental factors that could contribute to differences in QM between hospitals, interviewees mentioned legal frameworks, external pressures (e.g., inspections, thirdparty interests, shortage of nursing staff ), and cooperation among hospitals.As one interviewee noted: It also depends a lot on cooperation and collaboration [between hospitals]-whether one hospital is a driver for the other hospital or if joint projects or instruments are promoted.(Interview 2)

Discussion
We conducted a two-phase, sequential mixed-methods study using large-scale quantitative analyses followed by in-depth expert interviews to examine how organizational and environmental factors drive variation in the adoption of QM among hospitals.Our quantitative findings suggest that the hospitals that are more likely to adopt a greater number of QM practices are those that are larger in size, are located in big cities, have teaching hospital status, are less specialized, have a QM steering committee, possess an emergency department, are staffed with a high number of nursing personnel, and have a female quality manager.
Our results thus largely support the mechanisms expected based on institutional theory, resource dependence theory, contingency theory, and strategic choice theory.In short, it seems that hospitals are indeed adjusting their activities in response to external pressures, environmental conditions, and the requirements of resource providers, such as government and regulatory bodies, health insurance companies, medical suppliers and pharmaceutical companies, and skilled health care professionals.More broadly, our results support the theory-derived assumption that both organizational and environmental factors matter when it comes to the adoption of QM practices.This highlights the importance of considering both dimensions when designing and implementing QM approaches in hospitals.Neglecting either of these dimensions could result in a poor fit between the QM practices a hospital uses and the quality challenges it faces.
Most of the organizational factors we included in our analyses were significantly associated with the number of QM practices used by a hospital, collectively explaining large proportions of the variation between hospitals and supporting the majority of our hypotheses.For instance, we found that large hospitals used significantly more QM practices than small ones.Although our interviewees confirmed the theoretical argument that this difference probably relates to varying resource availability, they also provided many counterexamples, including the strategy of some small hospitals to use QM as a means of ensuring quality under conditions of scarcity.These insights might help explain contradictory findings in previous empirical studies (Hammer et al., 2013;Zarei et al., 2019).
Our findings also provide new insights into the relationship between QM-related organizational factors and QM variation, improving our understanding of what drives differences in the adoption of QM among hospitals.For example, although we found that the gender of the quality manager was significantly associated with the adoption of QM practices, many interviewees reported not observing gender differences in this regard.This suggests that gender does not have a direct, intuitive impact on QM adoption but rather might indirectly affect QM through other factors, such as communication or leadership style (Sabharwal et al., 2017).Furthermore, the disparity between this quantitative finding and the expectations of interviewees suggests that there might be a lack of awareness of gender as a source of systematic variation in the adoption of QM.
Our interviews provided valuable context to our quantitative findings regarding the impact of an emergency department on the adoption of QM practices.Although our quantitative finding was in concordance with our theoretical proposition (Hypothesis 6), the majority of interviewees expressed surprise at this.This discrepancy suggests that the influence on the adoption of QM practices might not stem directly from the presence of an emergency department.Instead, it might be the complexity of the organization or the urgency of cases seen by a hospital that drives the adoption of more QM practices, with the emergency department servicing as a proxy for these factors.Given the limited understanding of how hospitals with and without emergency departments differ in their operations and challenges, future research should attempt to identify any differences and their implications for QM.
Furthermore, our results suggest that environmental factors, despite being mostly beyond the control of hospitals, also contribute to variation in the adoption of QM practices.According to our results, factors that tend to be stable over time (i.e., hospital location by state and degree of urbanization) were significantly associated with hospitals' QM.This suggests that regulation may moderate differences between hospitals, for example, by aligning regulations across states to ensure that all states have the same requirements for QM or by supporting the implementation of QM in rural areas.Because the degree of urbanization was consistently linked to competition by our interviewees and most interviewees were surprised by our quantitative finding that competition was not significantly associated with the adoption of QM practices, further research is needed to investigate the mediating effect of hospital competition in the relationship between the degree of urbanization and hospitals' use of QM.
Our interviewees' perceptions of some factors, such as the presence of an emergency department and the degree of urbanization, suggest that our study was able to identify an underlying complexity in the relationship between organizational/ environmental factors and the adoption of QM practices that quantitative methodology alone could not capture.Indeed, the diversity of opinion we found among quality managers and the variety of explanations they offered for our quantitative findings (see Tables 4 and 5) suggest that the mechanisms that drive QM adoption cannot easily be reduced to numbers and objective measures.Overall, our study highlights the value of taking a holistic and multifaceted approach to studying the complex phenomenon of variation in the adoption of QM practices in hospitals.
Importantly, by highlighting potential contributors to variation in QM adoption beyond those examined in the quantitative phase of our study, our qualitative findings suggest avenues for further research.For example, interviewees indicated that beyond general organizational factors, factors that specifically relate to a hospital's QM department contribute to variation in QM-for example, the amount of financial or staff resources a hospital dedicates to QM.Although we included two such factors in our study-namely, the gender of the quality manager and the existence of a QM steering committee-these further suggestions by our interviewees point to other factors that future researchers may wish to explore.
Our study has a number of limitations that must be taken into account when interpreting its results.First, we used the hospitals' structured quality reports as our main data source in the quantitative phase of our study.Although these contain standardized information and are subject to the German Federal Joint Committee's specifications in terms of content and scope, they provide self-reported hospital information and do not allow for in-depth insights into the actual implementation of QM practices in hospitals in terms of scope and effectiveness.Second, regarding the dependent variable of our study, the total number of QM practices does not necessarily reflect a hospital's actual QM efforts.Further research is needed to determine whether a larger number of QM practices is in fact better in terms of process and quality outcomes.Third, it is important to note that the strength of the associations between the organizational and environmental factors and QM may depend on the specific context, and therefore, further research is needed to confirm these findings and identify the most effective strategies for broadening QM in different hospital settings.Fourth, while we analyze a comprehensive set of potential influencing factors, these explain only 25% of the variation in the number of QM practices between hospitals.Although this percentage is comparable to that in other studies using similar methods (e.g., Schneider et al., 2021), additional factors have been identified, both in our interviews and in the literature, as potential contributors to variation in the adoption of QM practices.Future researchers may wish to build upon our findings and identify other salient factors.

Practice Implications
The results of our study can be useful for informing the development of targeted interventions aimed at strengthening QM in hospitals, as well as for guiding future research on this topic.A better understanding of the factors that influence hospitals' adoption of QM practices can help policymakers identify hospitals that might need extra support to implement QM.Our results suggest, for example, that small rural hospitals are less likely to use a larger number of QM practices.Although small rural hospitals may have less need for QM practices (e.g., for structuring processes), the use of only very limited QM may not be sufficient to address all of the challenges they face in patient care and safety.Strategies to strengthen QM could involve providing incentives for these hospitals to become more involved in QM.Furthermore, a more targeted regulatory approach, informed by the distinct organizational and environmental characteristics of each hospital, may be beneficial (Jennings et al., 2019;Schneider et al., 2021).Such an approach would enable regulators to provide the necessary support for these hospitals to broaden their QM, tailored to their unique circumstances.Additionally, for benchmarking purposes, our results offer valuable insights for hospital management.Hospitals can either compare to hospitals that-according to the specific organizational and environmental features-are expected to be alike in their QM, or they can compare to hospitals that-according to the specific organizational and environmental features-are expected to be most comprehensive in their QM.Although the first approach might produce a more realistic benchmark achievable quickly, the latter might provide a longer term goal.Both approaches can be valuable for management to identify areas for improvement, devise targeted strategies, and set specific goals for improvement.
I don't have any budget at all.Others might have a QM budget.[…] Deciding whether I can use a patient survey tool or buy a piece of software-it's not something I can control.[…] So you end up being very dependent on what the managing director has in mind.You don't have much influence.(Interview 13) In addition, the degree to which hospital management and staff supported and accepted QM was mentioned: One aspect I also consider very important […] is the attitude of the management.[…] Without their support, QM becomes [merely] a default or compulsory program.[…] Another key factor alongside the management is the team, the employees […] all employees of the hospital […].The quality of cooperation [among staff] influences the extent of implementation and the integration of QM tools [into our practices].(Interview 2) […].If economic performance deteriorates, QM efforts are scaled down in the following years.(Interview16)

TABLE 1 :
Hypotheses regarding associations between specific organizational factors and variation in the adoption of quality management (QM) practices

TABLE 2 :
Hypotheses regarding associations between specific environmental factors and variation in the adoption of quality management (QM) practices

TABLE 3 :
Regression results on factors driving differences in the adoption of quality management practices among hospitals Note.n = 7,454 hospital-year observations from 1,571 individual hospitals.Between-effects coefficients are reported.Because of space limitations, single coefficients for the states are not reported in this table.Contact the authors for complete results.QM = quality management; bSE = bootstrapped standard errors.
Differences in Hospitals' Quality Management www.hcmrjournal.com

TABLE 4 :
Selected quotes regarding associations between organizational factors and the adoption of quality management (QM) practices

TABLE 4 :
Selected quotes regarding associations between organizational factors and the adoption of quality management (QM) practices, Continued

TABLE 5 :
Selected quotes regarding associations between environmental factors and the adoption of quality management (QM) practices