| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the 1Division of Research and the George Anderson Outcomes Unit, Department of Obstetrics and Gynecology, Women and Infants Hospital, Brown Medical School, Providence, Rhode Island.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: We performed a retrospective cohort study of 686 patients who underwent hysterectomy between January 1997 and September 1997 using medical chart review and hospital financial information. Demographic information, surgical approach (abdominal, vaginal, or laparoscopic), and surgical and postoperative factors were extracted from the medical record. Hospital charges were obtained from the hospital billing database. Relationships between charges and various clinical and demographic variables were examined using 
2, Fisher exact test, t tests, or analysis of variance, where appropriate. Logistic regression was used to estimate odds ratios while controlling for important confounding variables.
RESULTS: In our logistic regression model, blood loss greater than 1,000 mL (odds ratio [OR] 11.8, 95% confidence interval [CI] 4.2–33.2) and operative time 105 minutes or more (OR 14.2, CI 5.8–34.5) were strongly associated with higher charges for hysterectomy. Other factors associated with higher charges included: postoperative fever (OR 2.2, CI 1.1–4.5), increasing length of hospitalization (OR 5.3, CI 3.7–7.7), the use of prophylactic antibiotics (OR 3.0, CI 1.3–6.6), and the laparoscopic surgical approach compared with vaginal hysterectomy (OR 2.7, CI 1.0–7.0).
CONCLUSION: Surgical factors such as operative time and blood loss were strongly associated with increased hospital charges for hysterectomy.
LEVEL OF EVIDENCE: II-2
The costs and charges attributed to hysterectomy are highly variable in the published literature. However, evidence is available that costs and charges differ depending on surgical approach. Studies are difficult to compare because they include varying components of hospital charges to obtain their total hospital charge. Fairly consistent is the finding that vaginal hysterectomy, with its shorter operating times and hospital stays, is lower in costs compared with total abdominal hysterectomy (TAH) or laparoscopic-assisted vaginal hysterectomy (LAVH).4,5 Less clear is the economic comparison of LAVH and TAH, because several studies have produced conflicting data. Increased operating time and the use of disposable instruments has been implicated as an important factor that substantially drives up the cost of laparoscopic surgery.4,7
Although overall cost of hysterectomy has been studied as described above, few studies have evaluated factors associated with actual hospital charges. Any factor that complicates the surgical procedure or postoperative course, such as excessive intraoperative blood loss and febrile morbidity, can be expected to increase the costs associated with the procedure.10,11
The objective of this analysis was to evaluate factors that are associated with hospital charges for hysterectomy, with specific attention to differences based on surgical approach. To investigate this association, we performed a chart review supplemented with hospital financial information from computerized databases. Our 3 main hypotheses were 1) the abdominal surgical approach to hysterectomy is associated with increased hospital charges when compared with the vaginal or laparoscopic approach; 2) the least costly surgical approach of the 3 is vaginal hysterectomy; and 3) febrile morbidity is associated with increased costs.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To select the study sample, a computerized list of women who had hysterectomy for a benign condition between January 1997 and September 1997 at Women and Infants Hospital of Rhode Island was generated. Hysterectomies performed by the abdominal (TAH), laparoscopic-assisted vaginal (LAVH), or the vaginal approach were eligible. We used the following 3 corresponding International Classification of Diseases, Ninth Revision, Clinical Modification procedure codes to identify our study population: 68.3, 68.4, or 68.5. We excluded women who had either a preoperative or postoperative diagnosis of malignancy of the genital tract, or who had a supracervical hysterectomy. We selected our sample by requesting medical records of consecutive patients from the computerized list who met the inclusion and exclusion criteria. Information from medical record charts was then abstracted as described below.
Trained chart abstracters performed medical record chart reviews. Data collected included demographic and historical information (eg, age, race or ethnicity, marital status), insurance type (Medicaid, Medicare, self, or none compared with private or health maintenance organization), parity, and body mass index (weight in kilograms divided by height meters squared). Information regarding surgical approach (TAH, LAVH, or vaginal), use of prophylactic antibiotics (yes or no), estimated blood loss during surgery (in milliliters), and total operative time (in minutes) was extracted from the medical records. Data regarding length of hospitalization (in days) and postoperative complications, including febrile morbidity or evidence of infection, were noted.
The following study definitions were used in this analysis: Febrile morbidity was defined as a single temperature greater than 38.4°C or greater than 38.0°C on 2 occasions at least 6 hours apart (excluding the first 24 hours after the hysterectomy). Postoperative infection was defined as clinical infection of the wound, urinary tract, pelvis, or other site, or pneumonia, in the immediate postoperative period and was based on clinical assessment as recorded in the medical record. Postoperative complication included any complication occurring in the postoperative period (eg, febrile morbidity, infection, thrombosis, transfusion, etc).
Because actual costs are difficult to determine, hospital charges were used as a proxy for costs. Hospital charges (dependant variable) were recorded in dollars from the hospital billing database and recorded as a continuous variable. Medical record and billing data were linked by the patient's medical record number and date of discharge.
Once records were pulled, data were abstracted from the medical records onto standardized data collection form by 2 trained data abstractors. Data were verified for accuracy by the data management team and maintained in a confidential computerized database which was only accessible to study staff.
Data were analyzed using STATA 9.0 (StataCorp, College Station, TX). Relationships between cost and various clinical and demographic variables were examined using 2, Fisher exact test, t tests, or analysis of variance, where appropriate.
In preparation for multivariate analysis, we collapsed the charges into 2 categories, minimal to moderate (< 75th percentile) and high (
75th percentile). Body mass index (BMI) was categorized by commonly accepted categories of underweight and normal (BMI < 25 kg/m2), overweight (25–29 kg/m2), and obese (> 29 kg/m2). Operative time was grouped by quartiles, and the highest 2 quartiles were collapsed into a single category ( 105 minutes). Excessive blood loss was defined as greater than 1,000 mL.
We used logistic regression analysis to determine the odds of high charges (charges > 75th percentile) associated with each study variable. First, univariate logistic regression was used to obtain unadjusted (crude) odds ratios for individual variables. These variables were age, race, BMI, parity, length of stay, procedure type, blood loss, operative time, presence of infection, prophylactic antibiotic use, and presence of fever.
Next, we conducted multivariate logistic regression using variables associated with high charges in the univariate regression. Multivariate logistic regression allows for the evaluation of the effect of each independent variable on the dependent variable (charges) while also controlling for the effect on other independent variables. Infection rate and febrile morbidity were highly correlated, so in an effort to avoid multicollinearity only febrile morbidity (the more objectively determined variable) was retained.
The sample for this study was all procedures performed in the 9-month period. After setting 
= 0.05 and ß = 0.20 (power = 80%), we calculated that 426 cases were needed to detect a 10% difference in charges between the groups Thus, with 686 cases in the selected period, we had > 80% power.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Several clinical variables relevant to the surgical procedures are also displayed in Table 1. The duration of surgery, quantified in minutes, ranged from a low of 15 minutes to a high of 375 minutes, or over 6 hours. The infection rate across the groups was 9.3%, and most patients received prophylactic antibiotics (81.5%).
In Table 2, differences in key characteristics are shown by procedure. As stated above, TAHs were the most frequent of the 3 procedure types. Total charges, length of stay, estimated blood loss, and presence of febrile morbidity (all factors that could drive hospital costs and therefore charges) were also most associated with TAH (P < .001). The use of prophylactic antibiotics varied among the groups, being highest in the vaginal group, followed by LAVH and then TAH (P < .001). However, the overall infection rate did not differ between groups (P = .87). Operative time was noted to be higher for LAVH, but this difference was not statistically significantly different from other operative approaches (P = .13).
|
The total hospital charges were examined first by division into 3 categories: low (less than the 25th percentile), middle (26–75th percentile), and high charges (greater than the 75th percentile; Table 3). Increased BMI, increased length of stay, surgical type, increased blood loss, increasing operative time, infection, and prophylactic antibiotic use were associated with the 75th percentile category (P < .001 for all). The presence of fever was also more common in the high charges category (P < .001).
|
Univariate logistic regression analyses yielded unadjusted odds ratios (ORs) for high charges, as shown in Table 4. Several characteristics were found to be associated with high charges on univariate analysis: BMI 30 kg/m2 or more, increased length of stay, TAH, estimated blood loss more than 1,000 mL, operative time 105 minutes or more, infection, receiving prophylactic antibiotics, and presence of fever. In the calculation of ORs for procedure types, vaginal was considered the reference group for ease of interpretation of the OR estimates. Total abdominal hysterectomy was associated with high charges (OR 2.3, 95% confidence interval [CI] 1.5–3.6) when compared with vaginal. Odds ratios were also calculated with TAH as the reference group to better compare it to LAVH (not shown in table). This analysis showed that LAVH had a decreased odds of high charges (OR 0.5, 95% CI 0.3–0.9) when compared with TAH.
|
Variables found to be significant in the unadjusted analysis were then examined in a multivariate model, shown in Table 5. Blood loss more than 1,000 mL (OR 11.8, 95% CI 4.2–33.2) and operative time 105 minutes or more (OR 14.2, 95% CI 5.8–34.5) were strongly associated with higher charges for hysterectomy. Other factors associated with higher charges included length of hospitalization (OR 5.3, 95% CI 3.7–7.7), the use of prophylactic antibiotics (OR 3.0, 95% CI 1.3–6.6), and postoperative febrile morbidity (OR 2.2, 95% CI 1.1–4.5). Compared with vaginal, TAH was no longer associated with high charges. After controlling for other confounding variables, LAVH was associated with high charges (OR 2.7, 95% CI 1.0–7.0) when compared with vaginal.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We found that that surgical approach was associated with hospital charges. Vaginal hysterectomy, as hypothesized, was associated with the lowest charges. In our study, the association of the surgical approach with high charges differed between our unadjusted and our adjusted models. Not adjusting for other factors, the abdominal surgical approach to hysterectomy was associated with increased hospital charges when compared with the vaginal or laparoscopic approach. Conversely, when other factors were adjusted for, LAVH was most associated with increased hospital charges. Several previous studies have investigated the effect of surgical approach on cost of hysterectomy. Vaginal hysterectomy, with its shorter operating times and hospital stays, has been shown to be lower in cost compared with TAH or LAVH.4,5 Studies comparing the costs between LAVH and TAH have been inconsistent.
Our finding that the abdominal approach was associated with higher charges when compared with the LAVH, without adjusting for other factors, was consistent with the findings of the randomized trial by Raju et al.6 They found that although resource use related to the operating suite was higher for the LAVH group, overall costs were higher for the TAH group, secondary to longer lengths of stay. This also may be due to increased blood loss and increased febrile morbidity associated both with abdominal hysterectomy and with increased charges. Also, the use of disposable instruments has been implicated as an important factor associated with hospital charges in LAVH.4,7 At our institution, disposable instruments were not used during the study period. This may have contributed to the decreased charges of LAVH compared with other studies using disposable instruments for LAVH.
Other studies have found that the LAVH approach is associated with higher overall hospital charges or costs when compared with other surgical approaches.4,5 In a 1996 a retrospective chart review, Dorsey et al4 noted higher hospital charges and costs for the laparoscopic approach ($6,116) compared with the abdominal ($5,084) or vaginal approach ($4,221). Increased operative time and the use of disposable instruments were implicated in this association. This was consistent with our finding that, when controlling for other factors such as hospital length of stay, LAVH was the approach most associated with increased charges.
Additionally, in 2004 Lenihan et al5 found that hospital charges were highest for LAVH and lowest for vaginal. This study also looked at more "indirect" costs of hysterectomy; the investigators looked at time lost from work for the patients and their families after different types of hysterectomy. Although LAVH was associated with the highest hospital charges, it was also associated with the least amount of time out of work for patients and families, which presumably would translate into decreased costs for employers. In their randomized trial, Falcone et al8 also found a faster return to work for LAVH when compared with TAH.
Consistent with our stated hypothesis, febrile morbidity was associated with increased charges in both our crude and adjusted models. Among posthysterectomy morbid events, febrile morbidity is the most commonly reported event.10 The Collaborative Review of Sterilization study of 1,851 women undergoing elective hysterectomies for benign disease reported a 7.2% incidence of unexplained fever with vaginal hysterectomy as compared with 16.8% with abdominal hysterectomy.10 The abdominal surgical approach and excess blood loss have been identified as strong risk factors for febrile morbidity after hysterectomy.11 Given that postoperative fever often leads to increased hospital length of stay and increased diagnostic testing, it is not surprising that it is associated with higher charges.
Interestingly, 2 surgical factors were found to be most strongly associated with high charges. Increased estimated blood loss and operative time were associated with the greatest odds of high charges. Increased operative difficulty secondary to a variety of causes such as larger uteri and adhesions could lead to increased blood loss and operative time and therefore higher charges. Additionally, surgical technique and efficiency could also affect these factors. We did not collect data regarding the characteristics of the surgeons. Certain surgeons may have very low or very high rates of complications, and surgeon's efficiency in the operating room also varies considerably. It is possible that focusing more attention on technique and efficiency could affect these factors and thereby affect the overall cost of hysterectomy.
One unexpected finding was that prophylactic antibiotic administration was associated with increased hospital charges. This is counter-intuitive, in that decreased febrile morbidity is associated with decreased charges, and prophylactic antibiotics are known to decrease febrile morbidity. The most frequent practice for prophylactic antibiotic administration is to give the antibiotic within 2 hours of the surgical incision, and the current guideline at our institution is administration within 30 minutes of the incision. This analysis did not examine the timing of the prophylactic antibiotic. Although 81.5% of women in our cohort received prophylactic antibiotics, only 50% received preoperative prophylactic antibiotics.12,13 Given this difference, inappropriate "prophylactic" administration (eg, administration during the procedure for presumably more complicated cases or multidose administration after the procedure) could have led to this unexpected association with higher charges. It is also possible that more expensive antibiotics were used for prophylaxis in more complicated cases, explaining in part the association with increased charges.
There are a number of limitations to this research that deserve mention. Our analysis evaluated charges rather than true costs. Although charge data are easily accessible, it does not represent true expenditures or resource use. This study evaluates patients who underwent a hysterectomy at a large academic women's hospital in the northeastern United States. We may not be able to generalize these findings to patients undergoing surgery in community settings or in other geographic areas.
Finally, given that data for this analysis is from 1997, the findings should be validated in a more contemporary population of women undergoing hysterectomy. It is possible that with increased surgeon experience with the LAVH technique, complications and operative time have decreased, and therefore associated charges have decreased as well.
Cost containment is essential as we enter a new era of health care. This research evaluated some of the main factors responsible for the charges associated with hysterectomy. This study showed that surgical characteristics such as operative time and excess blood loss were strongly associated with increased hospital charges. Careful attention to surgical technique and operative efficiency may help control the costs associated with this common surgical procedure.
|
Footnotes |
|---|
doi:10.1097/01.AOG.0000209196.86946.81
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Keshavarz H, Hillis SD, Kieke BA, Marchbanks PA. Hysterectomy surveillance—United States, 1994–1999. MMWR CDC Surveill Summ 2002; 51(5505):1–8.
3. Johnson N, Barlow D, Lethaby A, Tavender E, Curr E, Garry R. Surgical Approach to hysterectomy for benign gynecological disease. (Cochrane Review). In: The Cochrane Library, Issue 1, 2005. Oxford: Update Software.
4. Dorsey JH, Holtz PM, Griffiths RI, McGrath MM, Steinberg EP. Costs and charges associated with three alternative techniques of hysterectomy. N Engl J Med 1996;335:476–82.
5. Lenihan JP Jr, Kovanda C, Cammarano C. Comparison of laparoscopic-assisted vaginal hysterectomy with traditional hysterectomy for cost-effectiveness to employers. Am J Obstet Gynecol 2004;190:1714–20.[Medline]
6. Raju KS, Auld BJ. A randomised prospective study of laparoscopic vaginal hysterectomy versus abdominal hysterectomy each with bilateral salpingo-oophorectomy. Br J Obstet Gynaecol 1994;101:1068–71.[Medline]
7. Summitt RL Jr, Stovall TG, Steege JF, Lipscomb GH. A multicenter randomized comparison of laparoscopically assisted vaginal hysterectomy and abdominal hysterectomy in abdominal hysterectomy candidates. Obstet Gynecol 1998;92:321–6.[Abstract]
8. Falcone T, Paraiso MF, Mascha E. Prospective randomized clinical trial of laparoscopically assisted vaginal hysterectomy versus total abdominal hysterectomy. Am J Obstet Gynecol 1999;180:955–62.[Medline]
9. van den Eeden SK, Glasser M, Mathias SD, Colwell HH, Pasta DJ, Kunz K. Quality of life, health care utilization, and costs among women undergoing hysterectomy in a managed-care setting. Am J Obstet Gynecol 1998;178:91–100.[Medline]
10. Dicker RC, Greenspan JR, Strauss LT, Cowart MR, Scally MJ, Peterson HB, et al. Complications of abdominal and vaginal hysterectomy among women of reproductive age in the United States. The Collaborative Review of Sterilization. Am J Obstet Gynecol 1982;144:841–8.[Medline]
11. Mittendorf R, Aronson MP, Berry RE, Williams MA, Kupelnick B, Klickstein A, et al. Avoiding serious infections associated with abdominal hysterectomy: a meta-analysis of antibiotic prophylaxis. Am J Obstet Gynecol 1993;169:1119–24.[Medline]
12. DiLuigi AJ, Peipert JF, Weitzen S, Jamshidi RM. Prophylactic antibiotic administration prior to hysterectomy: a quality improvement initiative. J Reprod Med 2004;49:949–54.[Medline]
13. Peipert JF, Weitzen S, Cruickshank C, Story E, Ethridge D, Lapane K. Risk factors for febrile morbidity after hysterectomy. Obstet Gynecol 2004;103:86–91.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
