To deepen our understanding of organizations’ heterogeneous responses to institutional demand, we develop a ‘relational complexity’ argument to highlight organizations’ diverse institutional linkages as another important source of practice variation. We argue that diverse relations between organizations and the institutional authority can filter distinct institutional pressures and expectations, shape organizational interpretations of environmental demands, and thus trigger heterogeneous organizational practices. We adopt this theoretical framework and distinguish two types of institutional linkages with the state to understand different adoption patterns in corporate social responsibility (CSR) in its early stage of diffusion in China. Based on a national survey dataset consisting of 1,268 firms, our analyses show that firms having a stronger bureaucratic linkage with the state tend to focus on more visible external-oriented CSR practices. In contrast, those firms forming a closer partnership with the state through political or semi-political associations are more likely to take more extensive adoptions by further developing internal CSR structures. This study enriches the institutional analyses by shifting our attention to the relational dynamics between organizations and institutional authority as a key source of practice variation. It also has important implications to the research and practices of CSR in emerging economies.

Introduction: Patients with mild traumatic brain injury (mTBI) often present to the emergency department (ED). Incorrect diagnosis may delay appropriate treatment and recommendations for these patients, prolonging recovery. Notable proportions of missed mTBI diagnosis have been documented in children and athletes, while diagnosis of mTBI has not been examined in the general adult population. Methods: A prospective cohort study was conducted in one academic (site 1) and two non-academic (sites 2 and 3) EDs in Edmonton, Canada. On-site research assistants enrolled adult (>17 years) patients presenting within 72 hours of the injury event with clinical signs of mTBI and Glasgow comma scale score ≥13. Patient demographics, injury characteristics, and ED flow information were collected by chart review. Physician-administered questionnaires and patient interviews documented the recommendations given by emergency physicians at discharge. Bi-variable comparisons are reported using Pearson’s chi-square tests, Student’s t-tests or Mann-Whitney tests, as appropriate. Multivariate analyses were performed using logistic regression methods. Results: Overall, 130/250 enrolled patients were female, and the median age was 35. Proportions of successfully diagnosed mTBI varied significantly across study sites (Site 1: 89%; Site 2: 73%, Site 3: 53%; p>0.001). Patients without a diagnosis were less likely to receive a recommendation to follow-up with their family physician (OR=0.08; 95% CI: 0.03, 0.21) or advice about return to work (OR=0.17; 95% CI: 0.08, 0.04) or physical activity (OR=0.08; 95% CI: 0.04, 0.17). Patients with missed diagnoses had longer ED stays (median=5.0 hours; IQR: 3.8, 7.0) compared with diagnosed mTBI patients (median=3.9 hours; IQR: 3.0, 5.3). In the adjusted model, patients presenting to non-academic centers had reduced likelihood of mTBI diagnosis (Site 2: OR=0.21; 95% CI: 0.08, 0.58; Site 3: OR=0.07; 95% CI: 0.02, 0.24). Conclusion: The diagnostic accuracy of physicians assessing patients presenting with symptoms of mTBIs to these three EDs is suboptimal. The rates of missed diagnosis vary among EDs and were associated with length of ED stay. Closer examination of institutional factors, including diagnosis processes and personnel factors such as physician training, is needed to identify effective strategies to heighten the awareness of mTBI presentations.

Introduction: Individual and institutional disparities in CT imaging rates for patients with head injuries have long been recognized, leading to the development of well-validated clinical decision rules designed to standardize clinical practice. To assess their impact on current practice, we sought to evaluate variation in CT imaging by emergency physicians for patients presenting with head injury across the province of Alberta. Methods: A unique data warehouse merging administrative, clinical, and imaging platforms for 11 Alberta emergency departments (EDs) was created. Unique identifiers were obtained for all emergency physicians who were included in this analysis if they evaluated in excess of ten ED patients presenting with a chief complaint of “head injury”. Patients with high triage acuity (CTAS 1) were excluded, as were patients who were admitted to hospital. Descriptive statistics were employed to describe variation between physicians and sites for a 24 month period from 2013-2015. Results: 311 emergency physicians treating 20,797 patient encounters for head injury were included. Overall a total of 8,245 head injury patients (40%) received one or more CT scans. Physician variation across the 11 sites ranged from 4% -100% of head injury patients receiving a CT. Within sites CT ordering between physicians varied from 9-fold (4% - 36%) at the lowest variation site, to more than 20-fold (4% - 90%) at the highest variation site. After removing the 5% lowest and highest ordering physicians, variation in ordering continued to range from 10% - 72%. No trends were observed across the two years examined. Conclusion: This is the largest study to date examining physician level variation in CT ordering practices for ED head injury patients. We have identified marked persistent practice variation despite the presence of well-validated clinical decision rules and a relatively low risk medicolegal environment. Variable risk tolerance and limited use of validated clinical decision rules are likely contributors making this area an ideal focus for targeted interventions to improve imaging appropriateness and reduce practice variation.

We challenge assumptions that have guided much research and policy aimed at understanding and reducing medical practice variation. Paramount is the focus on doctors as the cause of variation to the neglect of other possible influences. Some research literature suggests that patients, families, and the community context of practice may also influence treatment decisions. Failure to question present assumptions, despite weak evidence in support of them, may account for inability to explain persistent practice variation, develop appropriate implementable guidelines, or anticipate the effect on treatment decisions of greater patient involvement. In this paper, we discuss the weak response to the NIH Consensus Conference on early stage breast cancer because it may have reflected these problems. We urge a more complex and more empirical approach in explaining treatment choice and guidelines sensitive to the potential for value differences.

Objectives: Due to the discovery in the 1980s that blood transfusion can transmit HIV, there has been increased interest in technologies that reduce the amount of allogeneic blood used during and after surgery. These technologies include drugs (aprotinin, tranexamic acid, epsilon-aminocaproic acid, erythropoietin), devices (cell salvage), and techniques (acute hemodilution, predeposited autologous donation). The purpose of this study was to ascertain the degree of practice variation, if any, that exists for eight technologies in nine countries in orthopedic and cardiac surgery.

Methods: In each country, either all hospitals or a random sample of hospitals with medical/surgical beds were surveyed between 1995 and 1997. Two instruments were used. The first instrument was a postcard that asked recipients whether the technologies were currently being used in their hospital for orthopedic and/or cardiac surgery to reduce perioperative allogeneic transfusion. The second questionnaire elicited information regarding the degree of use both in qualitative and quantitative terms. Data were collected, entered, and analyzed in each country, with summary results submitted to the Canadian coordinating center on a standardized data collection form.

Results: Pharmaceuticals were generally used in a much smaller proportion of hospitals in orthopedic than in cardiac surgery. Aprotinin and tranexamic acid were the drugs most frequently used in cardiac surgery. Nonpharmacological technologies were used to a greater degree than drugs in orthopedic surgery, although there was wide variation among technologies and countries. Acute hemodilution and cell salvage were used in a greater proportion of hospitals for cardiac surgery than orthopedic surgery.

Conclusions: The results of this survey indicate that there is considerable practice variation in the use of technologies to minimize exposure to perioperative allogeneic transfusion within and between countries.