Financial impact of concurrent coding.

In this article ...

Discover how using an encoder during the anesthesia preoperative evaluation has the potential to enhance surgical facility revenue due to increased third-party reimbursement.

Hospitals are reimbursed for inpatient admissions according to the Inpatient Prospective Payment System (IPPS), which classifies patients by disease-related groupings (DRGs) of diagnoses. Specific weights assigned to each DRG determine the relative reimbursement for each of the discharge diagnoses based on projected resource utilization. (1)

These weights are summed and divided by the number of patients to determine the Case-Mix Index (CMI) of a patient population. The weights are affected by the patient's comorbid conditions; more complex or severe comorbid conditions typically require greater resources and thus result in higher reimbursements to hospitals.

In addition to determining reimbursement, the DRG, CMI, and related indices are used by various organizations to risk-adjust patients, compare hospital mortality rates, and generate performance "report cards." (2) Therefore, clinicians' documentation of comorbid disease in the medical record is essential to accurately risk-adjust patients to ensure appropriate payments and mortality estimates. Billing personnel may under-code comorbid conditions because of unclear, ambiguous, or difficult-to-locate documentation.

The anesthesia preoperative evaluation is a reliable and concise source of comorbid documentation for surgical patients and, when used, provides a more accurate and generally higher-risk DRG classification. (3)

We have previously illustrated that the preoperative assessment, with the use of encoder software significantly enhances the generation of the ROM and SOI. (4) The ability to comprehensively document comorbid conditions, even with a tool such as an encoder, requires additional time that is typically not available when clinicians are preparing patients for surgery.

To optimize documentation, reduce patient risks and cancelled surgeries, and improve efficiency on the day of surgery, complex patients would ideally be evaluated in a preoperative clinic that would allow for additional information gathering and documentation, and anesthesiologists have long advocated for patients with multiple medical problems to be seen in such a clinic. (5) However, current professional fee remuneration is inadequate to support staffing such clinics.

The purposes of this study were:

1. To analyze the effect on patients' DRG classifications when generating comorbid codes, comparing the use of a concurrent encoder with the use of documentation from the patient's medical chart to generate these codes.

2. To evaluate consequent financial reimbursement. We hypothesize that use of such coding software increases documentation of comorbid conditions, CMI, and financial reimbursement, thus providing financial justification for anesthesia preoperative evaluation.

Methods

We conducted a cross-sectional study to identify comorbid diseases in a convenience sample of 400 patients who had surgery at two academic medical centers (Johns Hopkins Medical Institutions, Baltimore, Maryland, and the Ochsner Clinic, New Orleans, Louisiana) between September 2004 and January 2005.

Study subjects were selected randomly from the daily operating room schedule who were:

* Older than 18 years of age

* Scheduled for inpatient admission following surgery

* Classified as having an American Society of Anesthesiologists (ASA) physical status score of II, III, or IV

All admissions received both standard retrospective coding based on the medical record and concurrent coding from the anesthesia preoperative assessment. It was essential that comorbid codes identified by the preoperative assessment were added to diagnosis codes generated by medical records since many diagnosis codes that are used to establish a DRG include diagnoses that occur postoperatively.

Twenty-three cases were excluded, as their discharge records were not available to medical records because the discharge abstracting was not completed by the end of the study or medical records could not provide their DRG classification. Approval was obtained from institutional review boards at both institutions.

All-patient refined--disease-related grouping

For the state of Maryland, inpatient reimbursement is regulated by the Health Services Cost Review Commission (HSCRC), and this organization adopted the all-patient refined-disease-related groupings (APR-DRG) system of defining DRGs, rather than the Centers for Medicare and Medicaid Services-disease-related grouping (CMS-DRG) system used by most other states. (6)

Developed by 3M Health Information Systems (Murray, UT), the APR-DRG grouping system provides two subcomponent indices--the Severity of Illness Index (SOII) and the Risk of Mortality Index (ROMI), which classify patients into one of four mutually exclusive risk categories; both indices are assigned to each patient. International Disease Classification, 9th Revision Clinical Modification (ICD-9-CM) diagnosis codes were entered into the 3M APRDRG software tool (called "Grouper") to calculate SOII and ROMI, which are calculated from the aggregation of comorbid conditions, principle diagnoses, and procedures performed.

Retrospective medical records coding

Comorbid conditions were abstracted from patient medical charts by trained personnel from the medical records department, including conditions that may have been identified postoperatively. These conditions were converted to ICD-9-CM diagnosis codes and assigned DRGs through the use of the APR-DRG Grouper.

Concurrent coding

An anesthesiologist conducted an independent assessment of comorbid conditions, with the aid of the encoder (DocuCode, Docusys, Atlanta, GA), which prompts the user for determination of clinical conditions as it assigns specific ICD-9-CM diagnosis codes.

Comorbid conditions were determined using the standard preoperative assessment form completed by the anesthesiologist assigned to the case and a review of data from the patient's preoperative chart, with attention particularly focused on abnormal laboratory values and diagnostic test results.

For each patient, the additional ICD-9-CM diagnosis codes generated were added to the ICD-9-CM codes obtained from the medical records department, and the resultant set of codes was processed with the APR-DRG Grouper to create the concurrent SOII and ROMI.

The additional codes were added to the codes that were abstracted by medical records for each patient, as many comorbid conditions that affect the DRG may have occurred postoperatively and would not have been accounted for during the concurrent coding process. All additional ICD-9-CM diagnosis codes were validated by the medical records department, and those found to be invalid were removed.

Case-Mix index

The primary outcome variable for this study was the CMI associated with each type of coding. The CMI is the average weight that is calculated by using the weight assigned to each DRG, which increases directly with higher SOII scores; the result is then divided by the number of patients in the group being analyzed.

The APR-DRG weights for Maryland typically range between 0.0677 for the simplest and 28.4704 for the most complex diagnosis and comorbid combination. Each specific APR-DRG weight is determined by the specific DRG and SOII calculated from the patient's comorbid conditions.

Therefore, under the APR-DRG system, each DRG is associated with four unique weights (one for each SOII). Consequently, hospitals may use CMI for comparative analyses of all patients or for more granular analysis of specific units, such as surgical patients.

Monetary reimbursement

The secondary outcome variable for this study was the potential difference in monetary reimbursement associated with concurrent coding with an encoder versus retrospective coding by medical records.

The potential monetary reimbursement was calculated by multiplying the projected revenue of the surgical patients (191) at Hopkins for fiscal year 2005 by the percent change in CMI of the study group. (It was not calculated for subjects from the Ochsner Clinic because CMI data on the entire surgical population were not available for comparison due to disruption of systems at Ochsner following Hurricane Katrina.)

In addition to information captured for SOII and ROMI, data regarding the patients' ages, gender, race, and type of surgery were also abstracted from the discharge record. Type of surgery was classified by major organ system, as defined by the procedure codes established at discharge by medical records.

Statistical analysis

Proportions, means, and standard deviations were used to describe subject demographic characteristics, comorbid conditions, SOII, and ROMI. Impact of the enhanced comorbid disease coding system was evaluated by percent of patients whose SOII increased by at least one category.

Mean SOII scores with and without the concurrent coding were compared with a matched t test. The numbers of patients who shifted one, two, or three SOII categories were presented as percents. A Wilcoxon signed-rank test was used to test for differences in the distribution of SOII. A chi-squared test was used to evaluate differences in proportions.

Results

Between September 2004 and February 2005, 400 subjects were enrolled. The majority of subjects were Caucasian; the average age was 59 years. Forty-four percent of the subjects were ASA Class II; 47 percent were ASA Class III. The three most common surgery types were gastrointestinal (30 percent), genitourinary (29 percent), and musculoskeletal (20 percent) (Table 1).

Table 1

Demographic Characteristics of Surgical Patients from
Johns Hopkins Medical Institutions and Ochsner Clinic Foundation

Demographic Characteristics   Proportion (n = 377)

Age (years [+ or -] SD)          59 [+ or -]15
Male                                       45%

Race

White                                      72%
Black                                      20%
Others                                      8%

ASA Classification

ASA II                                     44%
ASA III                                    47%
ASA IV                                      9%

Primary Procedure

Gastroenterology                           30%
Genitourinary                              29%
Cardiovascular                             10%
Musculoskeletal                            20%
Integumentary                               4%
ENT                                         3%
Respiratory                                 3%
Hematolymphatic                             1%
Endocrine                                   0%
Miscellaneous                               0%

From Stonemetz J, Pbam JQ, Marino RJ, Ulatowski JA,
Pronovost PJ. "Effect of concurrent computerized documentation
of comorbid conditions on the risk of mortality index."
J Clin Outcomes Manage 2007; 14(9):499-503, with permission.

Concurrent encoding increased the CMI for the study group at Hopkins by 4.7 percent (Table 2), which represents a potential increase in hospital reimbursement of $15.2 million for all surgical patients (Table 3). CMI charges are applied to all surgical patients at Hopkins, which is not the case for hospitals located in states other than Maryland.

Table 2

Case-Mix Index of Surgical Patients

             Group Encoded by    Group Encoded
              Medical Records    by Concurrent
                                     Coding       % Change  Difference

Institution       (95% CI)          (95%CI)                  (95% CI)

                 (n = 191)         (n = 191)

Hopkins      1.67 (1.52--1.83)  .75 (1.59--1.91)    4.7%       0.08
                                                            (0.02-0.13)

From Stonemetz J, Pham JC, Marino RJ, Ulatowski JA,
Pronovost PJ. "Effect of concurrent computerized documentation
of comorbid conditions on the risk of mortality index."
J Clin Outcomes Manage 2007; 14(9):499--503, with permission.

Table 3

Financial Impact of Applying the Increase in Case-Mix Index
Obtained from Concurrent Coding to all Surgical Charges for
FY 2005

Description (95% CI)                     Johns Hopkins Medical
                                             Institutions

Actual Case-mix Index                      2.02 [+ or -] 2.18
(2005) (17,649 cases)

Surgical Reimbursement FY 2005                   $399
(in millions) (17,649 cases)

Predicted reimbursement from               $332 ($302-$363)
retrospective medical coding

Predicted reimbursement                    $348 ($315-$380)
from concurrent coding

Predicted additional reimbursement from    $15.2 ($4.09-$26.3)
concurrent coding (in millions)

On average, medical record review identified 8.3 (95 percent CI, 7.8-8.8) comorbid conditions, and concurrent coding identified 13.6 (95 percent CI, 13.0-14.2), a difference of 5.3 (95 percent CI, 5.6-5.1) comorbid diseases per patient.

Concurrent coding was associated with increased likelihood of diagnosing anemia, hypertension, obesity, diabetes, and chronic obstructive pulmonary disease (Table 4). Mean SOII was 1.92 and ROMI was 1.44 on medical record review, and 2.23 and 1.68, respectively, with concurrent coding (Table 5).

Table 4

Comorbid Conditions of Study Subjects by Method of ICD-9 Coding

Comorbid Conditions          Medical Records  Concurrent     p value
                                 Coding         Coding
                                   (%)            (%)
                                (n = 377)     (n = 377)

Hypertension                       54             60          <0.01

Obesity                            14             35          <0.01

Anemia                             18             36          <0.01

Malignancy                         27             30          <0.01

Diabetes                           15             21          <0.01

COPD                               15             20          <0.01

Metastatic solid tumor              9              9           0.16

Prior myocardial infarction         5              9          <0.01

Diabetes with complications         2              4          <0.01

Renal disease                       3              3              -

Mild liver disease                  2              2              -

Severe liver disease                1              2           0.32

Dementia                            1              1           0.32

COPD, chronic obstructive pulmonary disease.

From Stonemetz J, Pham JC, Marino RJ, Ulatowski JA, Pronovost PJ.
"Effect of concurrent computerized documentation of comorbid
conditions on the risk of mortality index." J Clin Outcomes
Manage 2007; 14(9):499--503, with permission.

Table 5

Risk of Mortality and Severity of Illness Index of Study Subjects
by Method of ICD-9 Coding

               Medical Records Coding  Concurrent Coding  p value
                     (n = 377)             (n = 377)

SOI (mean SD)        1.92 (0.80)          2.23 (0.76)      <0.01

I                        33%                 15%

II                       44%                 51%

III                      19%                 29%

IV                        3%                  5%

ROM (mean SD)        1.44 (0.70)         1.68 (0.80)       <0.01

I                        66%                 50%

II                       26%                 34%

III                       5%                 13%

IV                        2%                  3%

SOI, Severity of Illness; ROM, Risk of Mortality. From Stonemetz J,
Pham JC, Marino RJ, Ulatowski JA, Pronovost PJ. "Effect of concurrent
computerized documentation of comorbid conditions on the risk of
mortality index." J Clin Outcomes Manage 2007; 14(9):499-503,
with permission

With concurrent coding, SOII and ROMI increased by 16 percent and 17 percent, respectively. In addition, SOII increased by one category in 27 percent of subjects, and ROMI increased by one category in 23 percent of subjects.

Discussion

In this study, we demonstrated that computerized concurrent coding increased the CMI by 4.7 percent, potentially resulting in an increase in surgical reimbursement for one academic institution of over $15 million for fiscal year 2005 had these changes been applied to all surgical patients.

As these potential increases in third-party reimbursement can be achieved only with significant involvement of time and energy by clinicians--primarily anesthesiologists conducting the preoperative evaluation--they may provide justification for staffing of a preoperative evaluation clinic with anesthesiologists.

In addition to improving patient outcomes, (7) such clinics could provide time to obtain a detailed medical history and to document comorbid conditions in patients scheduled for major surgical procedures.

The present study illustrates that, through a comprehensive analysis of patient comorbidity, the anesthesia preoperative assessment can:

* Help to improve comorbid documentation, directly improving the CMI and, therefore, hospital reimbursement:

* Provide more accurate estimates of predicted hospital mortality based on changes in CMI.

All inpatient reimbursement is defined according to the parameters of the IPPS, established by CMS in 1984, the essence of which is the use of the DRG classification for all discharges, such that each discharge is associated with a numerical code that represents all conditions and procedures typical for those diagnoses involved.

The basic CMS-DRG system further reports the presence or absence of complications or comorbidities. For a DRG that contains comorbidities, the allowable reimbursement is significantly higher due to the presumed increase in resource requirements.

As DRGs are assigned, they generate a relative weight that increases as the complexity and acuity of the discharge diagnosis increases. A CMI is calculated by adding all of the diagnoses of a specific DRG weight and dividing the sum by the number of discharges of the diagnosis.

Performing this calculation for the entire surgical case volume allows hospitals to analyze the CMI for surgical cases. This methodology is routinely employed by hospital administrators in comparing year-to-year variations in acuity and in performing comparative financial analysis between institutions.

CMI was used in this study to make the acuity comparison between groups of patients at both institutions, as well as to calculate the potential financial impact of using concurrent coding for all surgical patients at one institution.

Focus on documenting and coding for comorbid conditions increases dramatically with transition to a more granular DRG system. Referred to as Medical Severity DRG (MS-DRG), (8) this system increases the number of DRGs to allow for better representation of patient severity and eliminate certain comorbid conditions that are perceived to be acquired complications.

One of the new requirements is that hospitals accurately define comorbid conditions that are "present on admission" to differentiate them from conditions that arise as a result of the hospitalization. This new DRG system is a modification of the APR-DRG system that was developed by 3M and has been used extensively by all hospitals in Maryland since 2005. (6)

Compared to the CMS-DRG, which can only accommodate eight additional comorbid conditions, the APR-DRG classification system can accommodate up to 30. The addition of these comorbid conditions and the assignment to one of four SOI indices allows for more granular definition of the surgical risk classification.

The weights assigned to each SOII of each APR-DRG are determined by the HSCRC of Maryland and are factored by the previous year's health care costs. Each weight is multiplied by the base payment rate for individual hospitals to determine the revenue or payment received for each patient's discharge.

The national base payment rate for CMS is $4990.60; consequently, a DRG that has a weight of 1.00 would result in payment of this base payment rate. Typically, more calculations are used to arrive at the specific payment; however, a full discussion of them is outside the scope of this paper. For this study, we focused our attention exclusively on the APR-DRG, and the resultant changes apply only to the state of Maryland and may not be easily applied to other states.

Hospital coders in medical records departments routinely search a patient's record after discharge, assigning ICD-9 codes to each medical condition documented in the chart. A critical aspect of this process is that the specificity of the documentation completed by the physician determines the appropriate diagnosis code that may be used.

Medical records coders are not allowed to infer more complex or higher-acuity diagnosis codes. For example, if upon reviewing a chart, the medical records reviewer notices that a patient has low serum hemoglobin and the physician did not diagnose the patient as having anemia, the coders are not allowed to code anemia. Moreover, the use of arrows, as in [down arrow]Hb--typical in physicians' shorthand--is not considered adequate documentation.

Hospitals continually attempt to educate physicians, hoping to enhance their documentation to improve coding and, consequently, risk-adjusted mortality and reimbursement. Unfortunately, these efforts are generally not very fruitful; because enhanced coding benefits the hospital rather than the physicians, the physicians do not perceive its direct benefit.

The lack of incentives to code comorbid diseases is not limited to anesthesiologists. Surgeons frequently document the conditions of the organ system that justify the surgical procedure but do not comprehensively document other comorbid conditions. Conversely, anesthesiologists routinely generate a more comprehensive analysis of comorbid conditions, as these conditions frequently affect the choice and outcome of the anesthetic used.

Four methods are commonly used to enhance documentation of comorbid diseases and reimbursement:

1. Sequencing--choosing a different diagnosis as the primary diagnosis

2. Selection--selecting a completely new principle diagnosis

3. Specification--defining the primary diagnosis more exactly

4. Supplementation--adding more secondary diagnoses (9)

Gibby et al. maintain that the anesthesia preoperative assessment is a frequent source for improving DRG comorbid documentation through the use of the supplementation method. (3)

An encoder is a software tool that contains the entire library of ICD-9 diagnosis codes and has the ability to determine appropriate codes based on medical terminology. For example, the user may indicate hypertension as a medical condition; the encoder will request answers to clinical questions, each leading to further questions, until enough information has been gathered to assign the 5-digit, specific ICD-9 diagnosis code.

An encoder allows users to answer questions regarding medical condition without knowing how to code and without understanding which ICD-9 code ultimately is selected. Of particular importance, the answers to the questions can be concatenated to create documentation that is then included in the patient's record, providing the documentation necessary to substantiate the ICD-9 comorbid codes and the subsequent DRG classification.

However, specific coding rules must be followed for generation of hospital charges. For example, if the code for diabetes with complications is chosen, the addition of codes for additional comorbidities will be limited, despite their presence and documentation by the provider. Consequently, all comorbid conditions must be validated by medical records prior to inclusion into the DRG calculation. The use of an encoder primarily enhances comorbid documentation through supplementation and specification.

It is clear that it is not feasible for a physician to invest the additional time necessary to thoroughly document all of the comorbid conditions in the few minutes typically spent assessing each patient prior to surgery. It is practical, however, to evaluate higher-risk patients in preoperative clinics, which improve efficiencies through reduced delays and cancellations on the day of surgery. (10)

Despite studies that demonstrate the benefits of preoperative evaluation clinics, such as enhanced patient satisfaction, (11) reduced unnecessary testing and consultation, (12), (13) and decreased hospital length of stay, (7) very few hospitals have been willing to invest the resources necessary to establish and operate them.

Many hospital administrators feel that it is difficult to justify the expenditure of additional resources to operate such clinics, particularly if they believe that their surgical cases are currently being adequately managed and that the economic benefits are invisible.

In addition to the financial benefits, a preoperative assessment clinical may also improve patient safety. During root-cause analysis of anesthesia-specific sentinel events from 1994 through 2004, The Joint Commission reported that >60 percent of these sentinel events were related to inadequate preoperative assessments. (14) Preoperative evaluation of medical conditions has clearly been established as a necessary step to optimize surgical outcomes; (15) yet, most patients with complex medical problems who present for high-risk surgical procedures are not evaluated by the anesthesia team until the day of surgery.

Although preoperative risk factors have been demonstrated to be very effective in predicting hospital charges for surgical patients, (16) uncoordinated requests for preoperative consultationst have demonstrated little value in affecting any improvement in surgical outcomes. (17) A recent study illustrated that collation of missing documentation and preoperative evaluation of surgical patients provides an opportunity for more effective patient management. (18)

Intuitively, few would argue against preoperative evaluation of complex medical-risk patients, particularly those scheduled for intermediate- and high-risk surgical procedures. The real impediment to implementing preoperative evaluation clinics in both academic and community practice has been their lack of potential for revenue generation. (19)

In a business model that is driven by the ability to generate professional or facility fees as justification for new or additional services, these types of clinics are difficult to substantiate. We realize that, in most cases, the professional fees generated from preoperative evaluation clinics would not provide reimbursement adequate to staff them.

The documented improved efficiencies alone have not provided adequate justification to warrant a proliferation of these clinics, and clearly, not every patient needs to be seen in one. In our opinion, it is likely that these clinics are best reserved for those patients with a combination of high medical risk (ASA III and IV) and the likelihood of intermediate- and high-risk surgical procedures.

It is for these patients with multiple comorbidities that the additional documentation may improve patient safety and enhance hospital revenue. The challenge is to convince hospital leaders that the benefits are sufficient to provide support for these climes, as this study suggests that the financial benefits of opening a preoperative anesthesia clinic are substantial.

We recognize several limitations to our study. First, we sampled a relatively small number of patients from two academic medical centers. This limited our ability to generalize our results. Second, the coding of comorbid conditions in this study was done by the researchers. It is unknown whether these results would be replicated in routine practice. Finally, we modeled the financial implications of this study at one academic institution in Maryland that has a unique reimbursement model. Improvements in CMI in other states will vary.

Acknowledgement: Software used for this project was provided by DocuSys (Atlanta, GA).

References

(1.) CMS Fact Sheet regarding IPPS. overview.asp#TopOfPage. Accessed January 2008.

(2.) Averill RF, Goldfield N, Hughes JS, Muldoon J, Gay J, McCullough E, Bonazelli J, Mullin R: What are APR-DRGs? An Introduction to Severity of Illness and Risk of Mortality Adjustment Methodology. Salt Lake City, UT: 3M Health Information Systems, 2003

(3.) Gibby GL, Paulus DA, Sirota DJ, Treloar RW, Jackson KI, Gravenstein JS, vander Aa JJ: Computerized pre-anesthetic evaluation results in additional abstracted comorbidity diagnoses. J Clin Monit 1997; 13:35-41

(4.) Stonemetz J, Pham JC, Marino RJ, Ulatowski JA, Pronovost PJ: Effect of concurrent computerized documentation of comorbid conditions on the risk of mortality index. J Clin Outcomes Manage 2007;14(9):499-503

(5.) Parker BM, Tetzlaff JE, Litaker DL, Maurer WG: Redefining the preoperative evaluation process and the role of the anesthesiologist. J Clin Anesth 2000;12:350-6

(6.) Health Services Cost Review Commission (HSCRC). The Transition to APR-DRGs and Related Methodological Changes: Final Staff Recommendation for APR transition and methodology for FY06. June 1, 2005. http://www.hscrc.state.md.us/current_policy_papers/documents/fsr_apr_drg_transition.doc. Accessed 05/20/07

(7.) Halaszynski TM, Juda R, Silverman DG: Optimizing postoperative outcomes with efficient preoperative assessment and management. Crit Care Med 2004; 32(4 Suppl):S76-86

(8.) CMS Payment System Fact Sheet. http://www.cms.hhs.gov/MLNProducts/downloads/AcutePaymtSvsfctsht.pdf. Accessed January 2008

(9.) Cohen BB, Pokras R, Meads MS, Krushat WM: How will diagnosis-related groups affect epidemiologic research? Am J Epidemiol 1987; 126:1-9

(10.) Ferschl MB, Tung A, Sweitzer B, Huo D, Glick DB. Preoperative clinic visits reduce operating room cancellations and delays. Anesthesiology 2005; 103:855-9

(11.) Hepner DL, Bader AM, Hurwitz S, Gustafson M, Tsen LC: Patient satisfaction with preoperative assessment in a preoperative assessment testing clinic. Anesth Analg 2004; 98:1099-105.

(12.) Fischer SP: Development and effectiveness of an anesthesia preoperative evaluation clinic in a teaching hospital. Anesthesiology 1996; 85(1):196-206

(13.) Tsen LC, Segal S, Pothier M, Hartley LH, Bader AM: The effect of alterations in a preoperative assessment clinic on reducing the number and improving the yield of cardiology consultations. Anesth Analg 2002; 95(6):1563-8

(14.) Joint Commission Sentinel Events Statistics for Anesthesia. http://www.jointcommission.org/NR/rdonIyres/E27861E3-F238-441F-BBFD-50BD00835DD6/0/se_rc_anethesia_related.jpg. Accessed January 2008

(15.) Mukherjee D, Eagle KA: Perioperative cardiac assessment for noncardiac surgery: eight steps to the best possible outcome. Circulation 2003; 107:2771-4

(16.) Davenport DL, Henderson WG, Khuri SF, Mentzer RM, Jr.: Preoperative risk factors and surgical complexity are more predictive of costs than postoperative complications: A case study using the National Surgical Quality Improvement Program (NSQIP) database. Ann Surg 2005; 242(4):463-8

(17.) Katz RI, Cimino L, Vitkun SA: Preoperative medical consultations: Impact on perioperative management and surgical outcome. Can J Anaesth 2005; 52:697-702

(18.) Correll DJ, Bader AM, Hull MW, Hsu C, Tsen LC, Hepner DL: Value of preoperative clinic visits in identifying issues with potential impact on operating room efficiency. Anesthesiology 2006; 105(6):1254-9

(19.) Flowerdew RM: Preanesthetic evaluation in private practice. Anesthesiol Clin North America 2004; 22:141-53

By Jerry Stonemetz, MD, Julius Pham, MD, PhD, Robert Marino, MD, John Ulatowski, MD, PhD, MBA, Peter Pronovost, MD, PhD

Jerry Stonemetz MD

Clinical associate Billing & Compliance officer, Dept. of Anesthesia & Critical Care Medicine, Johns Hopkins Medical Institute

[e-mail]

[ILLUSTRATION OMITTED]

Julius Cuong Pham MD, PhD

Assistant professor, Department of Emergency Medicine, Department of Anesthesia Critical Care Medicine, Johns Hopkins University School of Medicine

[ILLUSTRATION OMITTED]

Robert J. Marino MD

Medical director Covington OSC, Department of Anesthesiology, Ochsner Health System

[ILLUSTRATION OMITTED]

John A. Ulatowski MD, PhD, MBA

Professor and Director, Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine

[e-mail]

[ILLUSTRATION OMITTED]

Peter Pronovost MD. PhD

Professor of anesthesiology and critical care medicine, Medical director Center for Innovations in Quality Patient Care, Johns Hopkins University

[ILLUSTRATION OMITTED]