Hum. Reprod. Advance Access published April 5, 2011 Human Reproduction, Vol.0, No.0 pp. 1– 7, 2011 doi:10.1093/humrep/der060

ORIGINAL ARTICLE Reproductive epidemiology

Body mass and risk of complications after hysterectomy on benign indications Merete Osler 1,*, Signe Daugbjerg1,2, Birgitte Lidegaard Frederiksen1, and Bent Ottesen 2 1

Research Center for Prevention and Health, Glostrup University Hospital, Nordre Ringvej, 2600 Glostrup, Denmark 2Department of Gynaecology, The Juliane Marie Center, University Hospital of Copenhagen, Rigshospitalet, Copenhagen 2100 Ø, Denmark *Correspondence address. Tel: +45-38633780; E-mail: [email protected]

Submitted on November 23, 2010; resubmitted on February 8, 2011; accepted on February 10, 2011

background: This study examines BMI in relation to risk of complications after hysterectomy on benign indications, and explores whether any associations vary by route of surgery.

methods: In this cohort study, we included data on health and lifestyle collected prospectively for all hysterectomy referrals for benign indications in Denmark from 2004 to 2009. Logistic regression was used to investigate relationship between BMI and complications reported at surgery or during the first 30 days after surgery.

results: Of the 20 353 women with complete data, 6.0% had a BMI , 20 kg/m2, 31.9% with BMI between 25 and 30 kg/m2 (classified as overweight) and 17.5% with a BMI ≥ 30 kg/m2 (categorized as obese). The overall rate of complications was 17.6%, with bleeding being the most common specific complication (6.8%). After adjustment for age, ethnicity, education, indication for surgery, uterus weight, use of prophylaxis, American Society of Anaesthesiologists classification, co-morbidity status and route of hysterectomy, obesity was associated with an increased risk of heavy bleeding during surgery [odds ratio (OR) ¼ 3.64 (2.90–4.56)], all bleeding complications [OR ¼ 1.27 (1.08 – 1.48)] and infection [OR ¼ 1.47 (1.23 –1.77)]. The risk of all bleeding complications [OR ¼ 1.48 (1.28–1.82)] and re-operation [OR ¼ 1.66 (1.26 –2.17)] were also increased among women with a BMI , 20. This U-shaped relation between BMI and bleeding, and the association between high BMI and infections were only seen for the abdominal route [abdominal hysterectomy (AH)]. The risk of infections was elevated among women with BMI,20 who underwent laparoscopic surgery [laparoscopic hysterectomy (LH)].

conclusions: Obesity increases the risks of bleeding and infections after AH. A BMI below 20 seems to increase the risks of bleeding and infection after AH and LH, respectively. Key words: body mass index / hysterectomy / cohort study / post-operative complications

Hysterectomy is the most common major gynaecological operation in the world. A high BMI is a risk factor for a number of diseases leading to hysterectomy such as fibroids and abnormal uterine bleeding (Laughlin et al., 2010). BMI also seems to be associated with the risk of complications after most elective surgical procedures (Choban and Flancbaum, 1997); however, the studies that have examined whether BMI is associated with risk of complications after hysterectomy have been inconsistent (Holub et al., 2001; Lo¨fgren et al., 2004; Rasmussen et al., 2004; Chopin et al., 2009). In addition, many focused on one specific route of surgery, and included only small and selected groups of patients. One of the largest, recent studies included 1460 patients who underwent laparoscopic hysterectomy (LH) for benign conditions, and showed that a BMI .

30 kg/m2 was not associated with an increased risk of peri- or postoperative complications, but a longer operating time was found for the obese women (Chopin et al., 2009). In a larger, nationwide Swedish study, with focus on post-operative infections, Lo¨fgren et al. (2004) found that high BMI was a risk factor in 3297 women undergoing elective abdominal hysterectomy (AH) or vaginal hysterectomy (VH). In the present study, we use prospectively collected data from a large national database to analyse whether BMI is associated with an increased risk of complications within 30 days after hysterectomy for benign indications. We also explore whether any risks associated with BMI differed in relation to whether the route of surgery was abdominal, vaginal or laparoscopic.

& The Author 2011. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: [email protected]

2

Materials and Methods We used data from the 22 185 women reported to the Danish Hysterectomy Database (DHD) between January 2004 through December 2008. The DHD database was established in 2003 to monitor and improve surgical quality of benign hysterectomy in Denmark. The database was opened for registration in October 2003. The data collection process in the DHD is based on a structured registration form, which is completed by the surgeon whenever a woman is referred for a hysterectomy on a benign indication. The registration form consists of three parts: page 1, which includes baseline information completed before surgery on patient’s diseases, lifestyle, height and weight and indications for surgery; page 2, which collects information on surgical procedures and possible prophylactic regiments immediately following surgery; page 3, which has information on post-operative hospitalization, complications and re-operations completed at discharge [the registration form (in Danish) can be found at www.kliniskedatabaser.dk/artikeldataVis.asp?id=30&m=2]. The data are reported to the National Danish Patient Register (LPR), using the unique person identification number for each citizen in Denmark, to link the information. The national response rate for DHD has been 88 – 100%, throughout 2004 –2008, when the LPR with all patient contact with clinical hospital departments was used as gold standard. The agreement between the reported data and the patient’ files in LPR has also been found to be high (82 –100%) (Hansen et al., 2008). The DHD is accredited by the Danish National Board of Health and therefore no ethics review board approval or patient’s consent is required. The socioeconomic data for each patient in the study population were derived by record linkage to the population-based Integrated Database for Labour Market Research (IDA) in Statistics Denmark. The data were analysed and reported in accordance with the STROBE statement.

Assessment of BMI Body height and weight, as reported to the surgeon at referral, were used for calculation of BMI as weight in kg divided by height in meters squared. Women with a BMI ,20 kg/m2 were considered to have low BMI, while overweight was defined as BMI ≥ 25 kg/m2, and obesity as a BMI ≥ 30 kg/m2.

Outcome variables Overall complication rate was estimated as at least one specific complication [bleeding, infection, organ lesion or other (those numbered 3 – 5 in the text below)] reported at surgery or during the first 30 days postoperatively. In addition, the following 10 specific indicators of complications were examined: (i) the amount of bleeding during surgery in ml, (ii) heavy bleeding (defined as 1000 ml bleeding or more during surgery), (iii) all bleeding complications (peri- and post-operative vaginalvault/wound/intraabdominal bleeding or hematoma), (iv) infections (urinary, wound or intraabdominal), (v) organ lesions (urinary tract or bowel), (vi) other complications (urinary retention, pain and thromboembolic events), (vii) operation time, (viii) stay in hospital for more than 5 days, (ix) re-operation or (x) re-hospitalization within 30 days after surgery.

Covariates The following variables, which have previously been shown to be associated with BMI and/or risk of complications (Choban and Flancbaum, 1997; Chopin et al., 2009), were considered as potential confounding or mediating factors: age, ethnicity, education, smoking, indication for surgery, uterus weight, use of antibiotics and thrombosis prophylaxis, co-morbidities (hypertension and diabetes) and American Society of

Osler et al.

Anaesthesiologists (ASA) score. The clinical data were used as reported in the registration form. Few women had ASA score 4 and consequently this category was combined with 3. Information on education was achieved from the variable ‘highest attained education’ (HFFSP) of Statistics Denmark. Level of education was grouped in three categories: ‘Basic education’corresponding to 7 – 9 years of obligatory schooling. ‘Medium education’ was defined as 9 – 12 years schooling, while ‘Higher education’ included all higher educations, approximating .12 years of education.

Statistical analysis Differences in the distribution of variables by BMI were analysed using the x 2 test. Logistic and linear regression models were used to examine the influence of BMI on the various complications. Statistical interaction between BMI and route of hysterectomy was examined by adding an interaction term in the logistic regression model. Models with and without the interaction term were compared using likelihood ratio tests. Possible clustering within hospital departments were accounted for using the cluster option in STATA. Of the 22 185 women who underwent hysterectomy, 1832 (7.9%) had incomplete data concerning their BMI (n ¼ 1 710) or other co-variables [ASA score (n ¼ 91), indication (n ¼ 9), ethnicity (n ¼ 18) and hypertension (n ¼ 4)]. The patients with incomplete data were older and had more missing information on other variables (data not shown); hence they were excluded from the subsequent analysis. In addition, information on smoking were missing for 1448 patients, and we performed the analysis presented with all cases included in the models and missing information coded as unknown. All data analyses were performed on anonymous data following the instructions from the Danish Data Inspection.

Results Of the 20 353 women with complete information on all clinical variables, 1236 (6.0%) had a BMI , 20 kg/m2, 6487 (31.9%) were classified as overweight and 3582 (17.5%) as obese. Table I shows the distribution in percent of the different indicators of complications in relation to BMI. The overall rate of complications was 17.6% with bleeding being the most common specific complication (6.8%). Those with low BMI (20.5%) and obese women (19.4%) seemed to have an increased risk of complications compared with women with normal (17.7%) or overweight (16.0%). The risk of heavy bleeding and infections were highest among obese women, while the risk of experiencing any bleeding complication seemed to be highest in women with a BMI ,20 and ≥ 30 kg/m2 (Table I). The total amount of bleeding during the operation and operating time increased with 9.9 (9.2 –10.8) ml or 1.3 (1.2–1.4) min, respectively, per unit increase in BMI. The percentage of re-operations was highest among women with low BMI. Table I also gives the distribution of the other patient characteristics in relation to BMI. In general, differences between the groups were small, but with the exception of use of antibiotic prophylaxis (P ¼ 0.08), BMI was associated with all other co-variables [age, ethnicity, education, smoking, indication for surgery, uterus weight, thrombosis prophylaxis, ASA classification, co-morbidities and route of hysterectomy (all P-values ,0.01)]. Thus, compared with women with normal weight, the obese women were more often of nonwestern origin, were less educated, non-smokers, had higher ASA scores and higher prevalence of co-morbidities. They were also more often operated on the indication of menorrhagia/metrorrhagia,

3

BMI and complications after hysterectomy

Table I Percentage [cumulative incidence (risk)] of complications within 30 days after hysterectomy and patient characteristics in relation to BMI. BMI (kg/m2)

.............................................................................................................................