Health Care Management Science

, Volume 11, Issue 3, pp 228–239

Operating room management and operating room productivity: the case of Germany

Authors

  • Maresi Berry
    • Klinik für Anästhesiologie und Operative Intensivmedizin, Universitätsklinikum Mannheim gGmbHUniversity of Heidelberg
    • Terry College of BusinessUniversity of Georgia
  • Alexander Schleppers
    • Klinik für Anästhesiologie und Operative Intensivmedizin, Universitätsklinikum Mannheim gGmbHUniversity of Heidelberg
Article

DOI: 10.1007/s10729-007-9042-7

Cite this article as:
Berry, M., Berry-Stölzle, T. & Schleppers, A. Health Care Manage Sci (2008) 11: 228. doi:10.1007/s10729-007-9042-7

Abstract

We examine operating room productivity on the example of hospitals in Germany with independent anesthesiology departments. Linked to anesthesiology group literature, we use the ln(Total Surgical Time/Total Anesthesiologists Salary) as a proxy for operating room productivity. We test the association between operating room productivity and different structural, organizational and management characteristics based on survey data from 87 hospitals. Our empirical analysis links improved operating room productivity to greater operating room capacity, appropriate scheduling behavior and management methods to realign interests. From this analysis, the enforcing jurisdiction and avoiding advance over-scheduling appear to be the implementable tools for improving operating room productivity.

Keywords

Operating roomOperating room managementOrganizational structureHealth care managersOperating room productivityAnesthesia group productivityFinancial incentives

1 Introduction

In this paper we examine operating room productivity. We are interested in whether the existence of an independent operating room manager as well as other characteristics of the operating room management function affects this productivity. Operating room management is a new field in Germany and has many local solutions. Before 2002, few hospitals had a formal operating room management. After 2002, hospital cost compensation changed from cost-plus to prospective payment, in an environment of cost-cutting. We use survey data from 87 German hospitals for the reference year 2002 to analyze the effect of organization and structure. Our analysis shows that while structural characteristics have an important influence on operating room productivity, the implementation of the elementary management functions planning and controlling are critical to efficient operating room performance.

The operating suite is one of the most expensive parts of the hospital. Operating room costs account for about 25% of total costs for an in house patient, but can make up the entire bill for an ambulatory surgical patient [1].These high costs per activity make the operating room an important consideration in hospital organization.

Our analysis of operating room productivity is closely related to the literature on the productivity of anesthesiology groups (see, e.g., [26]). Abouleish et al. [7] define the productivity of an anesthesiology group as (Amount of Clinical Care Provided)/(Anesthesia Staffing Cost). Anesthesiologists in Germany are employees of a hospital, only work for this one hospital. The continuous presence of one anesthesiologist for the entire surgery is a legal requirement. Compounded with an operating room financing structure which shares responsibility with surgical departments and a scheduling practice which depends on the anesthesiology department to manage schedule changes, we argue surgical time made available per unit of input is a good proxy for the hospital’s operating room productivity. More precisely, we measure operating room productivity as ln(Total Incision-Closing Time/Total Anesthesiologists Salary). We use surgical time from incision to closing instead of anesthesia time because it is not dependent on the characteristics of the operating room suite, and, is comparable across different hospitals.

Measuring and comparing the clinical productivity of medical groups is the basis to manage business operations in hospitals (see, e.g., [810]). We take this concept one step further and explicitly examine the effect of different structural elements of the operating room management function on the yearly operating room productivity in a cross sectional analysis of 87 German hospitals. This macro-economic comparison across an industry grounded in organizational theory is the main contribution of our study. While some of the measures used in our multivariate model may be specific to the German hospital setting, the general conclusions derived from our results are applicable to any other environment as well.

As McIntosh et al. [11] point out, anesthesia group productivity, operating room productivity and operating room efficiency are inextricably linked. Therefore, our analysis also provides some insight into how different organizational forms of the operating room management function within a hospital effect operating room efficiency. However, we do not directly measure operating room efficiency as defined by Dexter [1113].

The remainder of this paper is organized as follows. In Section 2, we provide an overview of the German hospital system, focusing on the organization of the operating room. In Section 3, we then develop ten hypotheses based on management theory, current hospital practice and defining characteristics of the hospitals themselves. Section 4 describes the survey data used and the analysis performed to test our hypothesis. In Section 5, we discuss our statistical results as well as the environment for operating room management illustrated by the analysis. The final section concludes with a short summary of the practical implications of our results for inclusion in operating room management planning.

2 The Case of Germany

The German hospital system is a state and federally guided system for the provision of in-patient health care services for the population. Each state is legislatively responsible for assuring that the medical needs of all its population are met. Hospital planning is a state obligation. Private hospital providers were few in 2002, making up less than 10% of hospital beds. German hospitals are grouped into six different levels of care with clearly defined medical and surgical departments and a pre-determined minimum spectrum of services at each of these levels. Level of care and number of medical/surgical departments increase from: Basic, Standard, Expanded, Maximum. The additional levels are university clinic and specialty clinic. University Clinic is a separate category generally with services similar to the larger and more diverse Maximum institutions. Specialty clinics are generally smaller clinics with a limited mandate, often Traumatology. A hospital at the Basic level of care is required to have Internal Medicine and General Surgery, while the addition of Obstetric/Gynecological services including a birth center is associated with Standard Care. Neurosurgery is not found at hospitals below Expanded care. Hospitals are designated for a certain level of care according to the planning goals and any changes in the department and spectrum of services must be approved, generally by both the corresponding state government oversight committee and the national medical insurance committee. Salaries and tariffs in Germany are regulated by a nationwide union system which sets the scales for all workers according to industry; the only regional difference is a defined factor discounting salaries in the old East German states in comparison to the old West German states. These two important factors, level of care guidelines and national salary scales make for a homogenous hospital landscape.

Operating room practices and procedures in Germany are coming under scrutiny by hospital managers as the cost pressure on the health care system continues to increase. In German hospitals, most hospitals have less than 400 beds and belong in a Standard level of care or lower [14], over 20% of hospitals have less than 50 beds. Hospital compensation was on a cost-plus basis, with separate funding for capital and operation expenses. In 2002, laws had just been adopted changing future hospital compensation to a prospective payment system. A prospective payment system compensates a provider with a fee determined by the patient’s ailment with a factor of co-morbidity. In a German hospital, especially in 2002, surgical procedures are performed by in-house staff on in-patients. Hospitals have traditionally been rigorously barred from performing most outpatient procedures as this was seen as impinging on the division of health care services in the in-patient and office-based care. A surgeon or anesthesiologist has either an office-based practice or an appointment at a hospital as a salaried employee, while this does not exclude moonlighting opportunities, these are extremely limited. Hospital surgeons are also responsible for running the surgical wards, performing such tasks as making rounds, drawing blood, talking with family members. All ambulatory surgery cases make up only 11% of an estimated 5 million hospital based surgical procedures [14]. Surgeries can be elective, they just require an overnight stay either before or after the surgery to be considered in-patient and therefore eligible for hospital service. Individual in-hospital operating rooms and suites are traditionally assigned to individual in-house surgical services, who are also involved in their design, maintenance and non-anesthesia staffing. This involvement includes financing major purchases as well as the routine expenses such as consumables, sterilization of re-useables. The exact structure of the departmental financing is left up to each hospital and traditionally depended highly on the negotiating power of current and past chiefs of the surgical and anesthesiology departments. In summary, going into 2002, the operating rooms were managed based on local custom with operating rooms “belonging” to a specific surgical service. This affects the scheduling of cases. Since the surgical services are involved in purchasing and staffing decisions for “their” operating rooms, this service reserves the entire day’s capacity up to the end of the day’s surgical program. Operating room schedules are released by the surgical services to the anesthesia department on a daily basis, usually the afternoon before. These include a mix of elective and urgent cases. We use the definition of urgent which includes cases that cannot wait 3 days [11]. On the day of surgery, this plan will be updated to include a mix of elective, urgent and emergency cases (those which have developed overnight, but could wait until morning). Surgical services can aggressively schedule operating rooms without staffing problems because both surgeons and most patients are in-house. It is expected that less urgent cases will automatically move to the next day in the event of over-scheduling. An attempt is also made to limit how often elective patients are rescheduled, as this can negatively affect patient satisfaction. Once a surgical service has completed a large portion of its schedule for the day and has not added any cases, any excess operating room capacity could then be anticipated. As both patients and surgeons are in-house, this is only possible when the surgical service has no one eligible to be operated on or no surgeon remains on the ward available to operate. At this point, it is often the senior attending anesthesiologist for the day shift has the option of making this time available for either an emergency which has arisen during the day which has not been able to be accommodated elsewhere, to transfer the staff to other operating rooms or to move surgeries from other departments into that operating room. A usual condition for moving an elective or urgent case is that the added procedure not incur overtime for that operating room. The success of this system depends on having a pool of patients and surgeons available to fill available slots. The cost of an additional day on the ward to make this flexibility possible had been accepted under the old cost-plus compensation system. The new, prospective payment system, with its set payment schedule, reassigns the cost of this extra day to the hospital. This is just one example of how the new compensation system for hospital care affects the incentives for operation room management. Any explicit cost measurement or control efforts on the part of hospitals and individual departments were local efforts based on expected future pressures and/or inherent good management by those individuals controlling the surgical and/or anesthesiology departments.

The staffing of surgical departments is beyond the scope of this article. Our proxy uses “incision-closing” time as an output, what the surgeons do with their available time, while greatly interesting to hospital administrators, is linked to other incentives in the health care system. For our analysis, the “unlimited” availability of surgeons due to the in-house surgical staffing system makes any special provision for a surgeon’s availability unnecessary. A number of studies [15, 16] address surgical efficiency directly and more completely than any examination of our data would make possible.

In the daily management of the operating rooms, there are several restrictions which differ from practices in the US and other countries. The first is the legal requirement that there must be at least one anesthesiologist present for the entire length of any anesthesia or stand-by service. Certified registered nurse anesthesiologists or their equivalents do not exist and anesthesia nursing staff is not permitted to either perform or monitor anesthesia procedures. Anesthesiologists in training (minimum of 5 years) are only allowed to perform operations under supervision. Any hospital with an anesthesiology department can train, placing the burden disproportionately on those hospitals with multiple operating rooms and surgical departments. In addition, pre-operative screening of patients is rigorously delegated to the anesthesiologists. As the anesthesiologist is a full-time hourly employee of the hospital. Medical departments traditionally provide their own administration, with senior physicians taking over administrative duties. It must be considered that a senior physician generally earns less than a senior hospital administrator, making this division of labor more cost-effective than in environments with other cost profiles.1

A number of practitioners in Germany have reviewed the operating room management organization at their affiliated hospitals. Geldner et al. [17] are one of many that suggest an OP manager, independent of the interacting departments, act as a coordinator and distributor of operating room resources is the optimal solution. In Brinkmann et al. [18] the dissatisfaction of operating room personnel in German hospitals with the lack of coordinated work flow should provide additional motivation for management improvement. Riedl [19] describes a three level system where the surgical departments do the planning for their operating room capacity, the requests are then compiled by a coordinating team which is responsible for assigning operating room capacity and ensuring the implementation of the requests. On an operational level, there has been a study by Hanss et al. looking at anesthesiology staffing levels using overlapping induction [20] or the introduction of internal transfer pricing for increasing operating room efficiency by Schuster et al. [21]. While this body of work indicates a strong interest in the implementation of operating room management, it is clear that there is no common consensus on what an operating room manager’s role and responsibilities are to be. The case of Germany is ideal for examining structural, organizational and management characteristics because the local solutions traditionally in place have begun to transition into new solutions. In 2002, there was sufficient interest to conduct this survey, but change had not infiltrated sufficiently to blur the homogeneity of the hospital industry. The variety of different combinations of these three types of characteristics makes an independent analysis possible.

3 Hypotheses

The goal of this article is to identify management and organizational factors contributing to increased operating room productivity. We develop ten hypotheses which can be grouped into two general groups. Hypotheses related to hospital characteristics and hypotheses related to the functions of the operating room management. The first group of hypotheses examines hospital attributes to determine if hospital classification has an influence on operating room productivity. The second group deals with basic functions of management as described in the integration of classical administrative theory by R.C. Davis—planning, organizing and controlling [22], at the organizational and at the workflow level.

3.1 Hypotheses related to hospital characteristics

Hypothesis 1. Operating room productivity increases with hospital size

Traditionally, economies of scale are expected to be an important factor in purchasing and management costs. This premise has led to hospital mergers, in Germany as well as in the US, particularly as private hospital providers enter the market. When contractual agreements prevent an actual physical merger, it is not unknown for management and purchasing structures to be developed to include all of the associated clinics, representing a de facto merger. This includes the division of specialties over adjacent, but previously independent clinics. It is assumed that in personnel planning, a larger company is better able to assure that an expensive employee, like an operating room nurse or anesthesiologist is optimally utilized.

Hypothesis 2. Operating room productivity decreases with the scope of the service the hospital offers

Diseconomies of scope can mitigate the importance of economies of scale, a medium sized hospital with many small departments can fail to build up a knowledge base for the efficient utilization of resources, personnel as well as facilities. In US studies, there has been some evidence that specialization, or lack of scope, leads to increased efficiency and, hence, increased productivity [23].

Hypothesis 3. Operating room productivity is higher for hospitals run by private corporations compared to those run by the public sector

New on the German hospital spectrum are the private providers, who have been especially active in the acquisition of small and medium size hospitals. While some of the larger hospitals have also been privatized, such the primary stakeholder in these transitions remains, in almost all cases, the community. As of yet, these purely private providers make up a small part of the total hospital landscape.

The hoped for advantages of privatization are the ability to change inefficient organizational structures and habits. Here the distinction between for-profit and not-for-profit does not play a role. A primary goal of privatization is to increase the importance of managing costs, allowing the community responsible for the hospital to control the costs without getting involved in the details of hospital management. It is expected that a for-profit group will have additional incentive to stay under the operating budget by a margin acceptable to stockholders.

3.2 Hypotheses related to management functions

3.2.1 Organizational structure of the operating room management

Hypothesis 4. There is a positive relationship between operating room productivity and the existence of an operating room manager

The existence of an operating room manager represents an important step in the organizational awareness that an operating room requires a different organizational structure than the traditional hospital departments. We expect that an operating room manager has as an interest the proper functioning of the operating rooms under his or her jurisdiction. An independent agent is a preferred method of coordinating interdepartmental activities, especially if the interests of these departments may conflict [17]. However, this hypothesis makes no assumptions about which responsibilities or control mechanisms are available to the operating room manager.

Hypothesis 5. Operating room productivity is higher in hospitals with a separate operating room department

An extension of Hypothesis 4, Hypothesis 5 takes the concept of an independent agent one step further. Not only exists a formal operating room Manager, the entire operating room administration is a separate entity and, hence, independent of the clinical departments.

Hypothesis 6. Operating room productivity is higher in hospitals where there is a contractual agreement among clinical departments for operating room usage

A contractual agreement is a working document for all departments who interact in the operating room functions. By institutionalizing the agreement, it is easier for participating departments to make the necessary changes in their own structures for a more efficient operating room utilization [17].

3.2.2 Relationships and workflow in the Operating Room

Hypothesis 7. Operating room productivity is higher in hospitals where most operating rooms are located next to each other in one central facility

The two common models for the organization of the operating room suite are a centralized operating room complex and the decentralized organization, where every surgical department maintains its own operating suite. While the applicability of decentralized facilities is dependent the size of the surgical departments, it is expected that a centralized operating suite would be able to take advantage of joint pre- and post-operating room areas as well as an overlapping personnel management. Personnel costs makes up a substantial portion of operating room costs and most non-surgical operating room personnel is not specific to the type of surgery being performed, a central operating room tract allows this expensive resource to be shifted quickly and easily from one room to another. Synergistic effects are also expected. Any operating room manager would also have an easier time overseeing immediate workflow process.

Hypothesis 8. Regular meetings among operating room users have a positive relationship to operating room productivity

Interaction between the participants is assumed to have a positive effect on the outcome. A venue for the exchange of information, viewpoints and experiences can be used not only to propagate the productivity concepts, but also to make adjustments to procedures in a timely manner. Empirical evidence exists that interdisciplinary meetings by hospital departments can measurably improve patient management [24]. We expect these positive management effects to be present for operating room productivity as well.

3.2.3 Planning and controlling function of the operating room management

Hypothesis 9. Advance planning of operating room schedules has a positive relationship to operating room productivity

Advance planning is one of the basic tenants of the scientific management theory, proposed by Frederick Taylor as early as 1895. Scientific management, which is distilled out of the engineering sciences, emphasizes the need to have work and the individual assignments planned at least 1 day in advance [25]. The creation of a weekly and daily schedule, with procedures and operating room assignments for personnel and patients can be seen as one application of scientific management in the operating room. The schedule is an implicit agreement how personnel and resources are to be distributed so as to maximize the utilization of these expensive and limited resources. The schedule documents the two key elements of a successful surgical management (1) the decision to perform the surgical procedure on the part of all decision-making parties, including any applicable third-party payers; and (2) the allocation of the pertinent resources in the form of facilities, personnel and, if necessary, specialized equipment [26].

Hypothesis 10. The implementation of disincentives for departments making overly frequent changes in the daily operating room schedule has a positive relationship to operating room productivity

With this hypothesis we attempt to capture the concept of consequences for breaking the contract, perceived or actual, which is implicit in the daily operating room planning. This hypothesis measures the ability of the arbitrators of operating room management, regardless of the structure used, to enforce their decisions and met out consequences or sanctions for unacceptable infractions. According to the behavioral theory of the firm (e.g., [2730]) sanctions are an important instrument motivating employees to act in the interest of the firm. An operating room manager should be able to judge whether the infractions are endemic to the department or are a result of poor organization on the part of the surgical department and react accordingly. Clearly an emergency surgery department will have more last minute schedule changes than an urology department. Sanctions can include decreased operating room time through decreased staffing of that services operating rooms, a later start to the first surgical procedure of the day to less preferential treatment for any available capacity in other operating rooms. Therefore, we expect sanctions to inappropriate changes to the operating room plan to increase operating room productivity.

4 Data and Methodology

4.1 Data

The data for this analysis includes structural data about hospital size and operating spectrum as well as financial and organizational (personnel and facility) information about the anesthesiology department. We used part of the data gathered by a large survey of the German anesthesiology professional organizations. The joint working group Anesthesia and Economics from the Deutsche Gesellschaft für Anästhesie (DGAI) and the Berufsverband Deutscher Anästhesisten (BDA) sent out this survey to all anesthesiology departments in German, 1381 in all, to collect structural information about each hospital and the relevant anesthesiology department. The survey was sent out in the spring of 2003, the expected return date was July 31, 2003 and no surveys were accepted after November 2003. The reference year was 2002. 320 surveys were returned, but not all of them were completely filled out. Especially the questions covering internal financial information were often left out. Some hospitals did not wish to provide this information, and others did not have an accounting system which allows the separation of the financial data into the necessary components. For our analysis we selected those respondents who provided all information necessary to calculate our efficiency measure and test the developed hypotheses. This procedure leaves us with 88 complete data sets. We dropped one data set from our sample since it includes cost information not consistent with the German salary structure. Table 1 defines the variables included in our analysis. In Table 2, the percentiles generated from the raw data are included.
Table 1

Variables measured for each hospital, using wording of the survey

Variable

Description

Provider/sponsor

Is the hospital provider a private corporation? Coded as 1 for Yes and 0 otherwise

Number of beds

What is the total number of beds in the hospital? Mark one of the following categories: less than 150 (coded as 1), 150–200 (coded as 2), 200–500 (coded as 3), 500–1000 (coded as 4), more than 1000 (coded as 5)

Number of beds for departmentj

Number of beds for each of the surgical departments in the hospital separately

Number of surgical departments

Total number of surgical departments in the hospital. Variable derived by counting the number of departments from the above question

Herfindahl index

The Herfindahl index is a measure of concentration of the services provided by a hospital. The variable is calculated according to the following formula: \({{\sum {a_i^2 } } \mathord{\left/ {\vphantom {{\sum {a_i^2 } } {\left( {\sum {a_i } } \right)}}} \right. \kern-\nulldelimiterspace} {\left( {\sum {a_i } } \right)}}^2 \), where ai represents the number of beds in the surgical departments i

Level of care

In Germany each hospital is granted the right to provide a pre-specified level of care. The variable is coded as 6 for a University Clinic, 5 for Maximum Care, 4 for Expanded Care, 3 for Standard Care, 2 for Basic Care, and 1 for a Specialty Clinic

Number of OR

How many operating rooms does the hospital have? Mark one of the following categories: 1–3 (coded as 1), 4–6 (coded as 2), 7–9 (coded as 3), 10–14 (coded as 4), 15–20 (coded as 5), more than 20 (coded as 6)

Total incision-closing time

Aggregated surgical time of all operations performed in the hospital within the reference year (in minutes). Surgical time is defined as the time from incision to closing. For the regression analysis, we standardize this variable by dividing the value for each hospital through the maximum value in the dataset

Total anesthesiologists salary

Total salary of all anesthesiologists employed by the hospital in the reference year

East/west

Does the hospital use the tariff structure for a western province of Germany (coded as 1) to pay its employees, or does it use the tariff structure for an eastern province (coded as 0)?

Intensive care

Does the reported total anesthesiologists salary only include the costs for servicing the operating rooms (coded as 0), or does it also include the costs for servicing the intensive care units (coded as 1)?

OR productivity

Calculated for each hospital as: ln(Total Incision-Closing Time/Total Anesthesiologists Salary)

OR manager

Does the hospital have an operating room manager? Self reported measure. Coded as 1 for Yes and 0 for No

OR

Do the operating room belong to the surgical departments individually (coded as 0), or is the operating room suite a separate department or profit center (coded as 1)

OR contract

Is there a contractual agreement between the surgical departments specifying how to interact in the operating room functions? Coded 1 for Yes and 0 for No

Centralized OR

Does the hospital have a centralized operating room? Coded as 1 for Yes and 0 for No

OR meetings

Are there regular operating room meetings at the hospital?

Self reported measure. Coded 1 for Yes and 0 for No

Weekly OR schedule

Is there a weekly operating room schedule at the hospital?

Self reported measure. Coded 1 for Yes and 0 for No

Daily OR schedule

Is there a daily operating room schedule at the hospital?

Self reported measure. Coded 1 for Yes and 0 for No

Consequences

Are there punitive consequences for a department making overly frequent changes in the daily operating room schedule (e.g., reduction of allocated operating room time)?

Open question. Self reported measure. Coded 1 for Yes and 0 for No

Table 2

Measured values of the variables defined in Table 1

Variable

%

Min

Percentiles

10th

25th

50th

75th

90th

max

Provider/sponsor

11.5

       

Number of beds

 

<150

150–200

200–500

200–500

500–1000

>1000

>1000

Number of surgical departments

 

1

2

4

5

8

12

14

Herfindahl index

 

0.092

0.129

0.211

0.310

0.411

0.675

1.000

Level of care

 

1

2

3

3

4

5

6

Number of OR

 

1–3

1–3

4–6

4–6

10–14

>20

>20

Total Incision-Closing Time (min)

 

71,093

123,825

187,242

336,840

618,014

1,471,020

1,979,580

Total anesthesiologists salary (€)

 

210,204

416,814

669,045

999,867

1,605,236

4,121,244

7,218,900

East/west

81.6

       

Intensive care

47.1

       

OR productivity

 

−1.983

−1.642

−1.269

−1.085

−0.823

−0.685

−0.257

OR manager

27.6

       

OR independent

42.5

       

OR contract

57.5

       

Centralized OR

86.2

       

OR meetings

59.8

       

Weekly OR schedule

41.4

       

Daily OR schedule

97.7

       

Consequences

13.8

       

For variables that are yes/no, the table shows the overall percentage of all “yes”-cases combined. For continuous and ordered categorical variables, the minimum and maximum values and percentiles are presented

Table 3 presents the structural characteristics of the hospitals in our sample. Comparing our sample set of 87 hospitals with the universe of all German hospitals we see that our sample is weighted towards larger hospitals. In 2002, 55% of all hospitals were under 200 beds, 25% of all hospitals has less than 100 beds. 30% of all hospitals only had one or two departments, neither of which were an independent anesthesiology department, excluding them from our survey. 15% of our sample was hospitals with less than 200 beds. However, 80% of the beds in Germany are in hospitals over 200 beds, a group well represented in our study [14]. One must also factor in that larger hospitals have a disproportionately larger number of operations, due to the larger number of surgical departments. It is also questionable how much operating room management is required for a small facility with a single operating room and a total of 2 surgeons. Therefore we argue that our data represents a typical operating room scenario at a hospital with multiple operating rooms and an independent anesthesiology department. However, we have to admit that from a strict statistical standpoint we cannot draw conclusions for the universe of all German hospitals based on the results derived from our dataset, since our dataset only includes 6.3% of all sent out surveys.
Table 3

Structural characteristics of hospitals in the sample

 

N

%

Number of beds

 <150

5

5.7

 150–200

8

9.2

 200–500

43

49.4

 500–1000

19

21.8

 >1000

12

13.8

 Total

87

100.0

Level of care

 Specialty clinic

7

8.0

 Basic

10

11.5

 Regular

34

39.1

 Central

17

19.5

 Maximum

13

14.9

 University clinic

6

6.9

 Total

87

100.0

Provider

 Religious

18

20.7

 Community

43

49.4

 University clinic

7

8.0

 Private

10

11.5

 Other

9

10.3

 Total

87

100.0

There are three different variables measuring the size of the hospital and its operations in our data set: the total number of beds in the hospital, the number of operating rooms and the aggregated incision-closing time of all operations performed in this hospital in the year 2002. Since our sample includes the number of beds of each surgical department, we can calculate two measures of concentration of the service provided by the hospital: the number of surgical departments and the Herfindahl index (see Table 2). Since every hospital in Germany is assigned a level of care which means it has to have specific departments with minimum capacities and services, we expect the size of a hospital and the scope of its operations to be highly positively correlated with its level of care. Table 4 shows the Spearman’s rho correlation coefficients as well as the Pearson correlation coefficients of the size and scope variables and the level of care. All correlation coefficients have values grater than 0.67 and are highly significant with p values smaller than 10−3. These correlations show that the structure of the hospitals in our dataset is homogenous, and, hence, our dataset includes the main characteristics of the German hospital landscape. The strength of the correlations of these six variables also implies that we should not include more than one of these variables in a regression model to avoid multi-collinearity.
Table 4

Correlation table of scale and scope parameters

https://static-content.springer.com/image/art%3A10.1007%2Fs10729-007-9042-7/MediaObjects/10729_2007_9042_Figa_HTML.gif

aSignificant at 1%

4.2 Methodology

Our measure of operating room productivity is derived from the literature on anesthesia group productivity. Dexter et al., Abouleish et al., and Barker [27], for example, measure anesthesia group productivity as (Amount of Clinical Care Provided)/(Anesthesia Staffing Cost). Since anesthesiologists in Germany are employees of a hospital, only work for this one hospital, and the law requires the presence of one anesthesiologist for every surgery, we argue that the productivity of a hospital’s anesthesiology group is a good proxy for the hospitals operating room productivity. Mores precisely, we measure operating room productivity as ln(Total Incision-Closing Time/Total Anesthesiologists Salary). We use surgical time from incision to closing instead of anesthesia time because it is not dependent on the characteristics of the operating room suite, and, hence, is comparable across different hospitals. Anesthesia time from intubation to extubation, for example, is different for hospitals with and hospitals without a holding area. In hospitals with a holding area immediately preceding the operation room the patients are usually intubated before entering the operating room, whereas in hospitals with a common holding area, patient preparation, takes place in the holding area, while the intubation takes place in the operating room itself. We calculate the productivity ratio using yearly aggregates, and then take the natural logarithm of this ratio to get a normally distributed variable which can be used as dependent variable in an ordinary least squares (OLS) regression model. The Kolmogorov–Smirnov test of our operating room productivity measure cannot reject the null-hypotheses of a normally distributed random variable (p = 0.2).

We use an OLS regression model to analyze the effects of different hospital characteristics, especially management practices and organizational structures on operating room productivity simultaneously. Using the variable definitions from Table 1, the specification of the OLS model is
$$\begin{array}{*{20}c} {{\text{OR Productivity = Total}}\,{\text{Incision - Cloning}}\,{\text{Time}}\,{\text{ + }}\,{{{\text{Provider}}} \mathord{\left/ {\vphantom {{{\text{Provider}}} {{\text{Sponsor}}}}} \right. \kern-\nulldelimiterspace} {{\text{Sponsor}}}}{\text{ }}} \\ {{\text{ + OR}}\,{\text{Manager + OR}}\,{\text{Independent + OR}}\,{\text{Contract}}} \\ {{\text{ + Centralized}}\,{\text{OR + OR}}\,{\text{Meetings + Weekly}}\,{\text{OR}}\,{\text{Program}}} \\ {{\text{ + Consequences}}\,{\text{ + }}{{{\text{East}}} \mathord{\left/ {\vphantom {{{\text{East}}} {{\text{West}}\,{\text{ + }}\,{\text{Intensive}}\,{\text{Care}}\,{\text{ + }}\,\varepsilon }}} \right. \kern-\nulldelimiterspace} {{\text{West}}\,{\text{ + }}\,{\text{Intensive}}\,{\text{Care}}\,{\text{ + }}\,\varepsilon }}} \\ \end{array} \,$$
(1)

Three more comments on our operating room productivity measure: First, using the aggregated yearly salary of all anesthesiologists as denominator, we have to control for the effects that Germany has two widely used salary structures, which are separated by a fixed percentage. These are linked geographically to the territory of the old Federal Republic of Germany and the old German Democratic Republic. We introduced the East/West dummy variable into our model to control for this difference in payrolls. Second, there is some lack of discrimination in hospital accounting practices. A number of hospitals do not differentiate between physicians’ costs for intensive care and anesthesia services. In the survey, hospitals were asked whether they discriminated between anesthesia and intensive care services. The scaling of intensive care with hospital size allows us to compensate for this lack of accounting discrimination by introducing the Intensive Care-Dummy as control variable.2 Third, the ratio (Total Incision-Closing Time)/(Total Anesthesiologists Salary) cannot take on values below zero, and is, hence, censored at zero. The logarithmic transform of this ratio, however, is not restricted to positive values avoiding a possible bias in an OLS regression. To check the robustness of the OLS results, we also perform a censored tobit regression [31] assuming the operating room productivity measure is left- and right-censored at its minimum and maximum values. An introductory description of the tobit model can be found in, for instance, Johnston and DiNardo, Kmenta, Long, and Maddala [3235].

To avoid multi-collinearity in the OLS estimations we include only one of the six correlated scale/scope variable in the model. Therefore hypotheses one and two cannot tested separately. We report only the results for the model using total incision-closure time as a scale variable in this paper. Similar results can be obtained for the other five variables. Since the dummy variable Daily OR Program is equal to one in almost all cases, it is positively correlated with the constant term in the regression, and causes multi-collinearity when included in the model. Thus, we excluded this variable from our final model.

5 Results and Discussion

The OLS estimation results as well as the censored tobit estimation results of Eq. 1 are reported in Table 5. Since the coefficients and significance levels of both statistical procedures differ only marginally, we conclude that the OLS results are free of a potential bias from a censored dependent variable. Therefore, we will focus on the OLS results in our discussion in the remainder of this section.
Table 5

Regression of operating room productivity on structural characteristics of the operating room management function and controls

Variable

OLS Regression

 

Censored Tobit Model

Intercept

−0.886a

 

−0.887a

(−6.77)

 

(−7.16)

Total Incision-Closing Time

0.401a

 

0.409a

(2.85)

 

(3.07)

Provider/sponsor

0.129

 

0.144

(1.18)

 

(1.38)

OR manager

−0.072

 

−0.076

(−0.92)

 

(−1.02)

OR as independent department

−0.011

 

−0.007

(−0.17)

 

(−0.11)

OR contract

0.002

 

0.005

(0.03)

 

(0.08)

Central OR tract

0.060

 

0.063

(0.60)

 

(0.66)

OR meetings

0.105

 

0.108

(1.42)

 

(1.54)

Weekly OR program

−0.120b

 

−0.125b

(−1.71)

 

(−1.87)

Consequences

0.182b

 

0.185b

(1.84)

 

(1.97)

East/west

−0.261a

 

−0.266a

(−2.88)

 

(−3.08)

Intensive care

−0.383a

 

−0.389a

(−5.41)

 

(−5.79)

R2

0.398

χ2 statistic

44.17

Adjusted R2

0.310

Pseudo R2

0.574

T-statistics are in parentheses below each coefficient.

aSignificant at 1%

bSignificant at 10%

The coefficient of Total Incision-Closing Time is positive and significant at the 1% level. This means that hospitals operating more tend to have a more efficient operating room management. Or in other words, we find strong evidence for economies of scale in operating room management which supports Hypothesis 1.

Diseconomies of scope, Hypothesis 2, cannot be shown due to the homogeneous nature of hospital organization in Germany. The correlation of scale and scope is too strong in the data. Since there are very few specialized clinics, it was not possible to make a significant separate analysis.

The Provider/Sponsor variable is positive as predicted by Hypothesis 3. The effect of privatization seems to have a positive effect on operating room productivity. However the estimated coefficient is not significant. Whether this is due to the relatively few private providers or the relative newness of private management is unclear.

The existence of an operating room manager cannot be shown to have any significant effect on operating room productivity. The coefficient of the OR Manager variable is negative, and not significant. Hypothesis 4 is therefore neither supported nor rejected by our data. While these results may, at first seem unexpected, there is an explanation for this lack of significance. Management capability depends not on having the correct title, but on having the correct function. On the examination of further survey questions, it is clear that most operating room managers had been in existence for less than one year at the beginning of 2002. Even the most optimistic prognosis cannot expect measurable financial results in such a short time frame. It is also conceivable that some hospitals have the function of an operating room manager filled, without the explicit title. An anesthesiologist chief who is able to, through personal and professional contacts, wield a greater than usual influence, could also positively affect operating room productivity. The operating room manager function can also be part of corporate culture. A hospital culture which places a high premium on efficient operating room routines, possibly combined with a facilities structure which allows for fast and efficient patient transport (late patient arrival is one of the most common given reasons for late operating room begin) can also have a higher than anticipated productivity.

The coefficient of the OR Independent variable is slightly negative and insignificant. Therefore Hypothesis 5 can neither be supported nor rejected. Two possible explanations why we do not find a significant relationship between the existence of a separate operating room department and operating room productivity are compatible with organizational management theory. The first is that the department has not been in existence long enough to have a positive effect on performance. The other possibility is that the department does not have the necessary jurisdiction, particularly budgetary, to have an effect on performance.

A contractual agreement cannot be shown to have any significant effect on operating room productivity. The coefficient of the OR Contract variable is near zero and not significant at the 10 percent level. Thus, Hypothesis 6 is neither supported nor rejected by our data. Similar to Hypothesis 5, without any knowledge of the content or any enforcement provisions of the contract, it is difficult to make any assumptions about the potential effectiveness of the contract.

The coefficient of the Centralized OR variable is positive as predicted by Hypothesis 7. However, the coefficient is not significant. Based on organizational theory, we would expect that centralizing all or some operating room facilities leads to some gains in productivity. But whether simply putting the operating rooms next to each other guarantees any synergistic effects is unclear and might explain why we are unable to find a significant relationship between the Centralized OR variable and the operating room productivity measure.

We are unable to show any statistically significant effect for regular operating room meetings on productivity. The coefficient of the OR Meetings variable is positive as predicted by Hypothesis 8, but insignificant. Similar to Hypothesis 7, the benefit of regular meetings depends on synergistic effects due to the coming together of individuals with complementary skill classifications. The management benefit of such meetings is unfortunately not clearly defined. The benefit of any meeting depends not only on the participants being present, but whether the meeting is organized effectively and what goals are met. A single survey question is a poor tool to judge meeting effectiveness.

The coefficient of the Weekly OR Program variable is negative and significant at the 10%. This evidence that a weekly schedule has a negative impact on operating room productivity contradicts Hypothesis 9. The negative effect of a weekly schedule can be explained by considering the case mix in a particular operating room or suite. While many operations in German hospitals are elective, a substantial number of operations are either emergency, meaning they must be carried out immediately, or urgent, meaning they must take place within 3 days. Operating room scheduling research has focused on the scheduling of elective cases (see, e.g., [11, 12, 36]). The research shows for a mixed case load, only advance scheduling for substantially less than capacity has any benefits [27, 35]. Practically all hospitals practice daily scheduling. In German hospitals, all cases for a particular surgical service use the same operating facilities, limiting planning flexibility. In addition, cases are routinely shifted to the next day due to routine over-scheduling of the daily program. These factors make weekly scheduling difficult and on any large scale unwieldy, as reflected in our results.

The attachment of sanctions to the unacceptable infractions of the OR schedule is the most influential non-structural parameter. The coefficient of the Consequences variable is positive and significant at the 10% level supporting Hypothesis 10. Sanctions are instrumental to enforcing an efficient workflow in the operating suite, regardless of the underlying management form. The positive effect of enforcement is not entirely unexpected, controlling is one of the primary management functions and enforcement is an extension of controlling. This supports our theory that there a many local solutions in effect in the hospitals, and that different management structures properly implemented may be appropriate for different hospitals.

We accept that the existence of certain parameters do not inherently lead to increased productivity. However, we clearly show that some characteristics of the operating room management function are associated with higher levels of operating room productivity in German hospitals. We find three characteristics which significantly effect operating room productivity. Increased hospital size, consequences for unacceptable schedule infractions and not having a weekly schedule are associated with improved performance. Hospitals cannot change size radically, especially not in the German hospital system, where each provincial government has an interest in the hospital system structure in its territory. From this analysis, the enforcement jurisdiction and avoiding advance over scheduling appear to be important tools for improving operating room productivity.

6 Conclusion

The focus of our research is to analyze the influence of specific aspects of the operating room management function in a hospital on the operating room productivity of this hospital. We use survey data from 87 German hospitals for the reference year 2002 and perform a cross sectional analysis. Our analysis shows that hospital size is the single largest predictor of operating room productivity. However, this information is not particularly helpful to the operating room manager. He or she cannot change the size of the hospital to increase operating room productivity. For the operating room manager, the reliance on structural characteristics to increase productivity is insufficient, implementing the basic management methods such as planning and controlling are critical to efficient operating room performance. For many institutions, this may require a paradigm shift. For the fiercely independent departments to work together in the operating room requires not just a retooling of current workflow patterns, but a restructuring of loyalties. The operating room workday is a busy, dynamic time, where emergencies are the order of business. We show that attempts to compensate by weekly scheduling leaves departments unable to react well to the changes inherent in the operating room workday. Daily scheduling was practiced by almost all responding hospitals. However, daily schedules are still subject to changes and modifications, especially near the end of the surgical working day. Our results indicate that it is not the cooperative environment which leads to greater productivity; consequences for inappropriate infractions help to align interests. We are able to show that the use of this management tool for the alignment of interests results in an improvement in performance regardless of the underlying managerial approach. Our analysis links improved operating room productivity to greater operating room capacity, appropriate scheduling behavior and management methods to realign interests.

What we focus on in our analysis is how productive the anaesthesiology departments are in the environment in which they find themselves, both hospital and national. This macro-economic comparison across an industry, based on organizational theory is the main contribution of our study. Let us now address the limitations of our results and sketch a route for further research. First, even though the characteristics of the 87 hospitals in our sample seem to be representative for a typical operating room scenario in a German hospital, from a strict statistical standpoint, we cannot draw conclusions for the universe of all German hospitals based on our empirical results, since our dataset only includes 6.3% of all sent out surveys. Second, we do not examine the micro-level workflow within each hospital. Thus, our results relate organizational components of the OR management function to OR productivity, but do not explain how these productivity gains are achieved. Exploring this link between different organizational elements and case scheduling practices in the daily work of an OR manager provides an interesting future research question.

Footnotes
1

We would like to thank one of the reviewers for pointing out that readers of scheduling and anesthesiology group literature might be interested in what incision-to-closing hours per day were. We calculated an average incision-to-closing per anesthesiologist on payroll of 2.075 hours per day. It must be remembered that this does not equate either to incision-to-closing time for an individual operating room, ASA billable hours or even the anesthesia time for an anesthesiologist assigned to an OR work slot for any given day. We calculated that, simply based on the inability to staff more than one OR simultaneously and the limitation of an 8 hour workday, a German anesthesiologist is only able to conduct 57% as much anesthesia per year as his or her counterpart in a country with certified registered nurse anesthesiologists or their equivalent. The additional burden of administrative duties and the pre-operative screening requirement will further decrease this percentage.

 
2

The assumption that costs for servicing the intensive care unit are a constant proportion of total anesthesiologists costs is justified by the legal and institutional environment in Germany. Hospital can neither freely choose the size of their intensive care unit nor the size of any other department or unit within the hospital. Every hospital in Germany has the duty to provide a certain level of care (Versorgungsauftrag). There are 6 different levels, and the definitions of these levels of care are very precise and explicitly address issues like which departments a hospital has to have and which size these departments are likely to have. Additionally, the compensation system in 2002 was structured such that any deviation from the specified level of care results in financial disadvantages. The correlation of different scale and scope parameters provided in Table 4 underline the static structure of the German hospital system and show that our data set is representative of the German environment. However, we acknowledge that this assumption is not appropriate for a US dataset, and, hence, repeating our analysis as described for US data would not lead to sound results.

 

Copyright information

© Springer Science+Business Media, LLC 2007