Skip to main content Accessibility help
Hostname: page-component-7ccbd9845f-mpxzb Total loading time: 1.626 Render date: 2023-01-28T01:47:54.566Z Has data issue: true Feature Flags: { "useRatesEcommerce": false } hasContentIssue true

Impact of body mass on hospital resource use in total hip arthroplasty

Published online by Cambridge University Press:  01 August 2009

John A Batsis*
Section of General Internal Medicine, Department of Medicine, Dartmouth-Hitchcock Medical Center, 1 Medical Center Drive, Lebanon, NH 03756, USA
James M Naessens
Division of Health Care Policy and Research, Mayo Clinic College of Medicine, Rochester, MN, USA
Mark T Keegan
Department of Anesthesia, Mayo Clinic College of Medicine, Rochester, MN, USA
Amy E Wagie
Division of Health Care Policy and Research, Mayo Clinic College of Medicine, Rochester, MN, USA
Paul M Huddleston
Department of Orthopedic Surgery, Mayo Clinic College of Medicine, Rochester, MN, USA
Jeanne M Huddleston
Division of Health Care Policy and Research, Mayo Clinic College of Medicine, Rochester, MN, USA Division of Hospital Medicine, Department of Medicine, Mayo Clinic College of Medicine, Rochester, MN, USA
*Corresponding author: Email
Rights & Permissions[Opens in a new window]



To determine the impact of BMI on post-operative outcomes and resource utilization following elective total hip arthroplasty (THA).


A retrospective cohort analysis on all primary elective THA patients between 1996 and 2004. Primary outcomes investigated using regression analyses included length of stay (LOS) and costs (US dollars).


Mayo Clinic Rochester, a tertiary care centre.


Patients were stratified by pre-operative BMI as normal (18·5–24·9 kg/m2), overweight (25·0–29·9 kg/m2), obese (30·0–34·9 kg/m2) and morbidly obese (≥35·0 kg/m2). Of 5642 patients, 1362 (24·1 %) patients had a normal BMI, 2146 (38·0 %) were overweight, 1342 (23·8 %) were obese and 792 (14·0 %) were morbidly obese.


Adjusted LOS was similar among normal (4·99 d), overweight (5·00 d), obese (5·02 d) and morbidly obese (5·17 d) patients (P = 0·20). Adjusted overall episode costs were no different (P = 0·23) between the groups of normal ($17 211), overweight ($17 462), obese ($17 195) and morbidly obese ($17 655) patients. Overall operative and anaesthesia costs were higher in the morbidly obese group ($5688) than in normal ($5553), overweight ($5549) and obese ($5593) patients (P = 0·03). Operating room costs were higher in morbidly obese patients ($3418) than in normal ($3276), overweight ($3291) and obese ($3340) patients (P < 0·001). Post-operative costs were no different (P = 0·30). Blood bank costs differed (P = 0·002) and were lower in the morbidly obese group ($180) compared with the other patient groups (P < 0·05). Other differences in costs were not significant. Morbidly obese patients were more likely to be transferred to a nursing home (24·1 %) than normal (18·4 %), overweight (17·9 %) or obese (16·0 %) patients (P = 0·001 each). There were no differences in the composite endpoint of 30 d mortality, re-admissions, re-operations or intensive care unit utilization.


BMI in patients undergoing primary elective THA did not impact LOS or overall institutional acute care costs, despite higher operative costs in morbidly obese patients. Obesity does not increase resource utilization for elective THA.

Research Paper
Copyright © The Authors 2009

The National Health and Nutrition Examination Surveys revealed that the prevalence of obesity in the US population has increased from 14·5 % in 1976–80(Reference Flegal, Carroll, Kuczmarski and Johnson1) to 32·2 % in 2003–4(Reference Ogden, Carroll, Curtin, McDowell, Tabak and Flegal2). This trend is likely to continue, as will the economic burden of obesity-related disorders(Reference Gregg, Cheng, Cadwell, Imperatore, Williams, Flegal, Narayan and Williamson3). Patients with moderate to severe obesity have higher health-care utilization and costs(Reference Finkelstein, Fiebelkorn and Wang4, Reference Quesenberry, Caan and Jacobson5) compared with other standardized body mass groups. The direct and indirect costs of obesity-related disorders in the USA are estimated to be between $US 52 and $US 95 billion annually(Reference Finkelstein, Fiebelkorn and Wang4, Reference Wolf and Colditz6).

The increase in the obesity epidemic parallels that of hip osteoarthritis. Twenty million adults in the USA are estimated to have osteoarthritis(Reference Lawrence, Helmick and Arnett7). The total societal cost of caring for patients with arthritis has burgeoned from $US 20 billion in 1997 to $US 128 billion in 2003(8, Reference Madhok, Lewallen, Wallrichs, Ilstrup, Kurland and Melton9). Although age is the most important risk factor(Reference Birrell, Johnell and Silman10), obesity is also associated with the development of symptomatic osteoarthritis of the hip(Reference Powell, Teichtahl, Wluka and Cicuttini11, Reference Wendelboe, Hegmann, Biggs, Cox, Portmann, Gildea, Gren and Lyon12).

Total hip arthroplasty (THA) for osteoarthritis improves mobility, relieves osteoarthritic pain and is cost-effective in ageing adults(Reference Chang, Pellisier and Hazen13). From a societal perspective, THA is one of the costliest inpatient procedures considering the sheer volume performed(Reference DeFrances and Hall14). In 2004 alone, there were 235 000 elective THA performed in the USA(Reference Kozak, DeFrances and Hall15). Few studies have examined the impact of body mass on resource utilization in the orthopaedic population(Reference Jibodh, Gurkan and Wenz16, Reference Epstein, Read and Hoefer17). With the increasing prevalence of both obesity and osteoarthritis, our aim was to determine if hospital costs and length of stay (LOS) are impacted by BMI.

Experimental methods

Study design and setting

A retrospective, observational cohort study was conducted on all elective unilateral THA patients admitted between 1 January 1996 and 31 December 2004. Thirteen different orthopaedic surgical teams, including an attending orthopaedic surgeon and rotating residents, performed all THA procedures at all one hospital. All surgeries were performed from Monday to Friday. Participating patients had authorized the use of their medical records for research purposes. The study was approved by the Institutional Review Board.

Data sources

All elective primary THA performed at our institution were identified (n 6009) using a previously described joint registry(Reference Berry, Kessler and Morrey18). We specifically did not include in our query any of the following groups of patients including: those requiring urgent, revision or bilateral arthroplasties; patients transferred or initially treated at outside institutions; or a primary surgical indication of trauma or septic arthritis.

We abstracted clinical and demographic data, including age, gender, height and weight at the time of surgery, surgical indication, use of cement, dates of admission, surgery, death, discharge and last follow-up date from the joint registry. Type of anaesthesia (general, regional, combined), American Society of Anaesthesiologists (ASA) physical status, and the time and date of admissions and discharges from the intensive care unit (ICU) were abstracted from departmental databases. Other patient characteristics were obtained using the Decision Support System (DSS) database (Eclipsys, Boca Raton, FL, USA). Admission times, LOS, dismissal time and date, discharge disposition, physical location, and individual and composite centre costs were also obtained. Patient expiration was confirmed using our electronic medical record, and verified using state and federal death registries. Major co-morbidities documented in this database were identified to compute a composite Charlson co-morbidity index(Reference Charlson, Pompei, Ales and MacKenzie19), as well as individual indicators.

Variables and definitions

BMI was calculated by dividing weight, in kilograms, by the square of height, measured in metres. All measurements were performed at the time of admission by trained nursing personnel. Our cohort was classified as follows: normal BMI (18·5–24·9 kg/m2), overweight (25·0–29·9 kg/m2), obese (30·0–34·9 kg/m2) and morbidly obese (≥35·0 kg/m2). We defined LOS as the number of days from the time of admission to the time of discharge. Economic analyses were conducted from the provider perspective at the patient service level and focused on direct costs of care associated with hospital and physician services during inpatient stay. Administrative data sources identified medical resource utilization and were valued using standard methods. Hospital-billed services were valued using departmental cost-to-charge ratios, while physician services were weighted by Medicare reimbursement rates. Accounting practices remained unchanged during the study duration. All costs presented have been adjusted to reflect 2005 constant US dollars(Reference Long, Bannon, Zietlow, Helgeson, Harmsen, Smith, Ilstrup, Baerga-Varela and Sarr20). Blood bank charges were defined as the costs of storage, processing and transfusion delivery. A re-admission was defined as any re-admission to our institution related to the primary surgery performed within 30 d of dismissal. An in-hospital death was defined as a death during the indexed initial surgical episode, while 30 d mortality was defined as a death within thirty days of the indexed surgery. Patient re-operations were defined as patients requiring further surgical intervention, following the initial surgery. A priori, a composite endpoint including re-admissions, re-operations and 30 d mortality was developed to account for the anticipated low numbers of these outcomes in this elective surgical population and is referred to as 30 d outcomes(Reference Fong Soohoo, Zingmond, Lieberman and Ko21).

Study cohort

We excluded 367 patients from the analysis for the following reasons: incomplete cost data (n 227); multiple joint replacements during the indexed hospitalization (n 3); no research authorization (n 58). Underweight patients (BMI ≤ 18·5 kg/m2) were excluded due to sample size and analysis considerations (n 79). Our final cohort consisted of 5642 patients. With the observed sample size and overall variability, our study had 80 % power to detect a difference between each group as small as 0·19 d and $620 in costs between the comparison and obese groups.

Statistical analysis

Primary outcomes included determination of LOS and resource utilization. Analysis of resource utilization focused on three approaches as outlined in Table 1. These approaches were selected specifically to examine resource utilization according to BMI on three separate elements of hospital flow. Statistical comparisons on baseline variables focused on a four-way comparison between normal, overweight, obese and morbidly obese groups (Table 2). A χ 2 test was used for baseline characteristics, including gender, co-morbid conditions, anaesthesia type, admitting diagnosis, use of cemented arthroplasties, 30 d re-admission rates and discharge locations. Unadjusted LOS, unadjusted costs, age, ICU days, re-operations and ASA functional class were tested using an ANOVA. Univariate costs are demonstrated in Table 3. Bonferroni adjustments were performed where appropriate.

Table 1 Characteristics of three analyses examining resource utilization within the indexed surgical episode in patients undergoing total hip arthroplasty

PACU, post-anaesthesia care unit; PT, physical therapy; OT, occupational therapy; RT, respiratory therapy; E&M, evaluation and management.

Table 2 Characteristics of 5642 patients undergoing elective hip arthroplasty by BMI categoryFootnote *

ASA, American Society of Anaesthesiologists.

* All data are represented as means and counts, with their standard deviation or proportions in adjacent column.

Patient characteristics contain empty or too few data to perform statistical analyses.

Table 3 Univariate unadjusted length of stay and cost centre estimates of 5642 elective total hip arthroplasty patients by BMI categoryFootnote *

ICU, intensive care unit; PT, physical therapy; OT, occupational therapy; RT, respiratory therapy; E&M, evaluation and management.

* All cost data are represented as unadjusted costs and are rounded to the nearest dollar. P value <0·05 is considered statistically significant.

Multiple comparison analyses performed for individual comparisons between groups when ANOVA significant: a, normal BMI v. overweight; b, normal v. obese; c, normal v. morbidly obese; d, overweight v. obese; e, overweight v. morbidly obese; f, obese v. morbidly obese.

The effect of BMI on the primary outcomes of LOS, overall, hospital and physician costs was examined using the entire cohort after adjusting for the baseline and surgical covariates through use of generalized linear regression models, and testing for appropriate interactions (Table 4). In order to provide costs adjusted for covariates, separate regression models were analysed for each categorization of cost. Covariates included in this model were age, sex, ASA class, surgical indication, use of a cemented arthroplasty, admission the day before surgery, anaesthesia type and ICU stay. Because year of surgery and Charlson co-morbidities could affect our primary outcomes, these variables were also included. For adjustment, age was classified into five different categories: <55, 55–64, 65–69, 70–74 and >75 years. The 65–69 years age group was identified as the median age, and labelled the reference group. Each co-morbid condition was treated as an indicator variable. BMI was also treated as a nominal variable with four levels. Indicator variables were also assigned for the calendar year in which the patient was operated with 2004 as the reference category. We also adjusted for nursing unit as it has previously been demonstrated that it may impact costs(Reference Batsis, Naessens, Keegan, Huddleston, Wagie and Huddleston22).

Table 4 Multivariate analysis of length of stay and cost centre estimates of 5642 elective total hip arthroplasty patients by BMI categoryFootnote *

ICU, intensive care unit; PT, physical therapy; OT, occupational therapy; RT, respiratory therapy; E&M, evaluation and management.

* All cost data are represented as costs adjusted for age, sex, ASA (American Society of Anaesthesiologists) class, surgical indication, use of a cemented arthroplasty, admission the day before surgery, anaesthesia type, ICU stay, year of surgery, Charlson co-morbidities and nursing unit, and are rounded to the nearest dollar. P value <0·05 is considered statistically significant.

Multiple comparison analyses performed for individual comparisons between groups when ANOVA significant: a, normal BMI v. overweight; b, normal v. obese; c, normal v. morbidly obese; d, overweight v. obese; e, overweight v. morbidly obese; f, obese v. morbidly obese. P < 0·05 after Bonferroni adjustment.

Secondary outcomes included unexpected admission to the ICU from regular nursing floor, 30 d outcomes and discharge disposition. Assessments of the effects of BMI on transfer to the ICU, and on the combined endpoint of 30 d mortality, re-admissions or re-operations, were performed using logistic regression models adjusting for the specified covariates described above. A P value <0·05 was considered statistically significant. Analyses were performed using the SAS statistical software package version 9·1 (SAS Institute, Cary, NC, USA).


Patient baseline characteristics are outlined in Table 2. Age differed between BMI groups (P < 0·001) as did gender (P < 0·001). A higher BMI was associated with a higher ASA class (P < 0·001) and a higher overall Charlson co-morbidity index (P < 0·001).

Univariate costs are demonstrated in Table 3. Adjusted LOS was similar among normal (4·99 (sd 0·11) d), overweight (5·00 (sd 0·09) d), obese (5·02 (sd 0·11) d) and morbidly obese (5·17 (sd 0·14) d) patients (P = 0·20; Table 4). We analysed LOS after excluding nursing home patients but it was no different (P = 0·56; data not shown). Overall, hospital and physician costs in Analysis 1 were similar among all groups (P = 0·20, P = 0·25 and P = 0·19, respectively). Blood bank costs were significantly lower in morbidly obese patients ($180) than in obese ($206), overweight ($255) or normal BMI ($274) patients (ANOVA, P = 0·002; multiple comparison analysis, all P < 0·05). Post-operative overall, hospital and physician costs (Analysis 2) were no different between groups (P = 0·30, P = 0·32 and P = 0·24, respectively). Analysis 3 represents operative and anaesthesia costs, which were higher in the morbidly obese group than in all others (all P < 0·05). All other adjusted costs were non-significant and are represented in Table 4.

There were no statistical differences in our composite 30 d endpoint among the normal, overweight, obese or morbidly obese groups (6·7 % v. 6·2 % v. 6·6 % v. 7·6 %; P = 0·62). In addition, morbidly obese patients were more likely to be discharged to an assisted living facility or a nursing home (24·1 %) compared with normal BMI (18·4 %; P = 0·001), overweight (17·9 %; P < 0·001) or obese patients (16·0 %; P < 0·001). There were no differences in the number of patients transferred to the ICU (P = 0·47) or the number of ICU days (P = 0·88).


The relationship between BMI and health-care utilization is of vital importance due to the continual increase in the prevalence of obesity in the USA. Our key findings in this large inclusive cohort of over 5600 patients demonstrate that neither obese nor morbidly obese patients undergoing primary elective THA have differing LOS or overall hospital resource costs than normal BMI or overweight patients. Such findings are important in dispelling myths regarding the possible added resource use of obese patients in the peri-operative setting. However, our study is restricted to the hospital stay and lacks longer-term clinical outcome data and costs associated with the post-hospital course.

Length of stay is often a major determinant in overall inpatient medical costs. There was a lack of statistical or clinical significance between groups. A previous study demonstrated that LOS is related to BMI with a ‘J’- or ‘U’-shaped relationship in a US population cohort(Reference Zizza, Herring, Stevens and Popkin23); however, that particular study focused on all hospitalizations, including urgent surgeries which often portend to higher costs and may indeed be related to sicker patients, contrary to what is observed in an elective arthroplasty population. Thompson et al. estimated that future health-care costs will be higher for obese patients(Reference Thompson, Brown, Nichols, Elmer and Oster24) along with a need for arthroplasty, thereby making costs likely contributing factors. However, our results suggest that in-hospital costs for obese patients are no different, suggesting that much of the possible burden, in attempts of reducing LOS, have been shifted to rehabilitation centres or home-health services, entities not captured by our data set.

Our study results confirm the results of Jibodh et al., who did not observe any differences in resource utilization among the four BMI categories they examined in elective THA patients(Reference Jibodh, Gurkan and Wenz16). The costs outlined in their paper ranged from $13 355 in the group with BMI < 25 kg/m2 to $14 055 in the morbidly obese group, all of which are markedly lower than what we have demonstrated in the current study. Whether their annualized costs were standardized as ours were, is unknown. Furthermore, that paper also did not observe any differences in LOS. The only other known study examining this relationship was conducted by Epstein et al.(Reference Epstein, Read and Hoefer17) in 1987; however, their data can be considered out-of-date not only due to practice changes, differing surgical techniques and changes in interdisciplinary care, but also because their study was in an era where LOS was markedly longer than what is expected nowadays.

The observation of lower blood bank costs in the morbidly obese group is inconsistent with previous data which did not demonstrate any cost differences(Reference Jibodh, Gurkan and Wenz16) or an association between obesity and increased intra-operative blood loss(Reference Lehman, Capello and Feinberg25, Reference Bowditch and Villar26). As administrative data sets were used in our study, we were unable to determine the reasons for laboratory tests or the particular indications for transfusion. One possibility may be that morbidly obese patients have higher blood volumes, and can therefore lose more blood before their Hb may reach a level where transfusion is indicated. In addition, anaesthesia and operating room costs (Analysis 3, Table 4) were statistically higher in morbidly obese patients than other BMI categories, suggesting that there were other miscellaneous factors involved that were not captured by our databases. These higher costs, though, are consistent with the higher operative and anaesthesia times observed in our cohort.

Although we hypothesized that obese or morbidly obese patients would have a higher number of unexpected ICU admissions and ICU days, no differences were observed. The relationship between obesity and resource utilization has been examined in detail in the ICU setting and following cardiac surgery(Reference Schwann, Habib, Zacharias, Parenteau, Riordan, Durham and Engoren27, Reference Finkielman, Gajic and Afessa28). Studies have been equivocal, with some demonstrating a ‘U’-shaped relationship between costs and BMI, but others failing to find such a relationship(Reference Ray, Matchett, Baker, Wasser and Young29). Although we did not specifically measure ICU complications, our results may be due to the fact that patients admitted for elective surgical intervention have pre-operative assessments whose primary purpose is to optimize their surgical candidacy and identify peri-operative strategies to reduce surgical and post-operative complications.

We used a composite 30 d endpoint for death, re-admissions and re-operations, as we had insufficient power to examine these outcomes individually. Obese patients have a higher peri-operative mortality risk due to their coexisting co-morbidities(Reference Prem, Mensheha and McKelvey30). Because our centre also acts as a referral centre, we likely would be unable to capture all relevant clinical data (reason for re-admission and reason for re-operation at the patient’s local facility) if a standard 90 d post-operative payment period was used; hence our data may underestimate the impact of BMI on longer-term outcomes. Yet, the present data are consistent with studies that have not shown any identifiable increases in mortality or post-operative complications among obese or morbidly obese patients undergoing cholecystectomy, hysterectomy or even cardiac surgery(Reference Prem, Mensheha and McKelvey30, Reference Moulton, Creswell, Mackey, Cox and Rosenbloom31). Although there was an increasing non-statistical trend in this outcome as BMI increased, it is unknown as to whether this would differ at a higher BMI. We agree, though, that obesity is likely not associated with higher peri-operative risk or an increased risk in treatment(Reference Jiganti, Goldstein and Williams32).

Obesity is known to be a risk factor for long-term nursing home placement(Reference Valiyeva, Russell, Miller and Safford33). There are some studies which have demonstrated that orthopaedic surgery predicts nursing home placement(Reference Legner, Doerner, Reilly and McCormick34), but to the best of our knowledge, no data exist which primarily examine impact of obesity in post-operative elective hip arthroplasty on nursing home placement, particularly with regard to short-term placement. Our results demonstrate that morbidly obese patients were more likely to be discharged to a care facility compared with the other BMI categories following elective arthroplasty.

Peri-operative complications and care of the surgical obese patient are of concern, both to medical staff and hospital administrators. The risks associated with surgery in this population may lead to increased costs, particularly in the morbidly obese group, as has been demonstrated in our study. This analysis focuses on an economically important issue from an institutional perspective, as to whether obese or morbidly obese patients, populations at higher risk for requiring arthroplasties, have higher resource use following elective primary THA. Although the costs differences per patient may not seem overtly large, when one considers that over 200 000 hip arthroplasties are performed annually, one cannot ignore the overall economic and societal impact, particularly in a health-care system such as that in the USA where the burden of costs rests upon public funding in this population group(Reference Lubitz and Riley35, Reference Waldo, Sonnefeld, McKusick and Arnett36).

The study results can be applied to high-volume tertiary care centres with orthopaedic expertise. Such centres have lower lengths of stay and costs than corresponding lower-volume centres(Reference Martineau, Filion, Huk, Zukor, Eisenberg and Antoniou37). Previous population-based epidemiological studies demonstrated that our institution’s patient population, which is of white Northern European descent in >90 % of cases, often is generalizable to a significant portion of the US white population(Reference Melton38). However, we caution that our results may not necessarily be extrapolated to other races or ethnic backgrounds and that further studies are warranted to better elucidate our hypothesis in these patients(39). By using standardized measurements, we eliminated measurement bias. Many studies estimate BMI using self-reported height and weight, which generally underestimate BMI(Reference Rowland40). Our costs were also standardized and adjusted for inflation. Finally, we limited our registry search to focus on unilateral elective primary THA patients to minimize any potential confounding, as urgent repairs, revisions, septic arthritis and bilateral procedures are associated with higher costs, lengths of stay and complication rates(Reference Bozic, Katz, Cisternas, Ono, Ries and Showstack41). Our results can therefore provide excellent external validity only to such patients.

Our study has the inherent limitations of a retrospective study. We relied on the validity of the data in various administrative databases, whose use has been challenged by others(Reference Iezzoni, Foley, Daley, Hughes, Fisher and Heeren42, Reference Mears, Bawa, Pietryak, Jones, Rajadhyaksha, Hungerford and Mont43). Minor procedures and diagnoses may be ignored in chart abstraction or incorrectly entered; however, the likelihood that these differences exist should be similar between groups. Administrative databases do not permit abstraction of the time and date when all hospital staff deemed that the patient was ready for discharge. Hence we relied on the actual time of discharge, which can be heavily dependent on availability at skilled nursing facilities. However, after excluding patients discharged to such facilities, our results were no different. Furthermore, a 90 d period in ascertaining costs and outcomes would be more appropriate, but could not be performed using our data set.

Our results may underestimate the differences in LOS and costs observed in obese or morbidly obese patients. Although we adjusted for baseline characteristics and medical co-morbidities, including diabetes and cardiac disease which are known to portend to higher costs(Reference Hogan, Dall and Nikolov44, Reference Nichols and Brown45), patients with notable co-morbid conditions may not have been surgical candidates, thereby introducing an element of selection bias. It is possible that some of the underlying differences across weight groups were removed through regression covariates. Well-known variables, including ASA, Charlson co-morbidity, surgical year and indication, cemented arthroplasty, admission the day prior to surgery, anaesthesia type and ICU stay, all are known to impact resource utility. If differences on these variables are due to BMI, the adjusted analysis will underestimate any cost differences. In fact, although there was a higher proportion of patients with degenerative joint disease in the morbidly obese group, patients with this indication had mean costs $1002 lower than other surgical indications. Additionally, there were higher proportions of obese and morbidly obese patients in years since 2000, but the costs of more recent years were $2319 less expensive. These results suggest that, despite baseline factors that were more prevalent in the obese and morbidly obese patient groups, the analysis appropriately adjusted for these variables to isolate the impact of BMI on costs. We relied on the definitions of obesity recommended by the WHO, as opposed to those based on life-insurance tables or older methodologies. Previous studies have determined that obese patients have 50 % higher health-care costs than ‘normal’ weight patients; hence we wished to determine the costs in accordance to this BMI category(Reference Andreyeva, Sturm and Ringel46).

Although BMI is a widely used surrogate for the degree of adiposity, it not only overestimates adiposity in patients with increased musculature, but more importantly underestimates adiposity amongst the elderly, who have age-related reductions in lean mass(Reference Marzetti and Leeuwenburgh47, Reference Gallagher, Visser, Sepulveda, Pierson, Harris and Heymsfield48). Recent data demonstrate the lower correlation coefficients between BMI and lean mass, and between BMI and percentage body fat, in the 65+ years age group compared with patients below the age of 65 years(Reference Smalley, Knerr, Kendrick, Colliver and Owen49). Hence we may have underestimated the true impact of adiposity on resource utilization. Finally, our analysis was limited to the inpatient stay and thus we cannot determine resource utilization past dismissal. As much of the costs are shifted to post-acute care venues, rehabilitation, home-health services and skilled nursing facility costs are factors which contribute to overall resource utilization, and are not reflected in our results, leading our results to have underestimated the impact of obesity on overall resource use(Reference Vincent, Weng and Vincent50). This study was undertaken from a hospital perspective as opposed to a societal perspective, thereby omitting potential non-medical costs which are incurred by patients and others related to the surgical intervention.

Our results cannot be extrapolated to patients with a BMI < 18·5 kg/m2 as we had excluded this subgroup of patients from this analysis. Furthermore, indications for arthroplasty are usually different in this cohort as they are less likely to develop osteoarthritis, can be considered frail and thus may not either be candidates for the intervention or may not obtain its benefits. In addition, the inherent selection bias by both primary care providers in referring these patients for surgery and orthopaedic surgeons in operating on them is real. The scope of these issues would need to be fully ascertained to better understand them.

Our findings have significant implications for physicians, policy makers and others interested in the impact of obesity in the hospitalized patient. Using this large cohort of patients, our data suggest that there is no difference in LOS or in overall costs. Identification of measures to optimize health-care delivery in obese patients may reduce inpatient costs but would likely impact all groups alike. Little attention has been given towards the ‘ageing obese’ population, the need for arthroplasty and its fiscal consequences. Prospective evaluation of the impact of obesity on patient-related outcomes and costs are needed to ascertain areas where clinical pathways or interventions can be implemented to sustain efficiency and fiscal responsibility.


The authors obtained funding through the Small Grants Program, Mayo Clinic Rochester. No authors have any conflicts of interests. The work was performed primarily while J.A.B. was at Mayo Clinic Rochester. J.A.B. was involved in conceptualization and design of the study, analysis and interpretation of the data, manuscript writing and approval of the final submitted manuscript. J.M.N. and A.E.W. were involved in the conceptualization and design of the study, analysis and interpretation of the data, statistical expertise, critical revisions of the manuscript, and approved the final submitted version. M.T.K. and P.M.H. were involved in conceptualization and design of the study, analysis and interpretation of the data, critical revision of the manuscript, and approved the final submitted version. J.M.H. was involved in conceptualization and design of the study, analysis and interpretation of the data, supervision, responsible for obtaining funding, and critically reviewed the manuscript, also approved the final submitted version. We thank Donna K. Lawson, LPN, for her assistance in data collection and management.


1.Flegal, KM, Carroll, MD, Kuczmarski, RJ & Johnson, CL (1998) Overweight and obesity in the United States: prevalence and trends, 1960–1994. Int J Obes Relat Metab Disord 22, 3947.CrossRefGoogle Scholar
2.Ogden, CL, Carroll, MD, Curtin, LR, McDowell, MA, Tabak, CJ & Flegal, KM (2006) Prevalence of overweight and obesity in the United States, 1999–2004. JAMA 295, 15491555.CrossRefGoogle ScholarPubMed
3.Gregg, E, Cheng, Y, Cadwell, B, Imperatore, G, Williams, D, Flegal, K, Narayan, KMV & Williamson, DF (2005) Secular trends in cardiovascular disease risk factors according to body mass index in US adults. JAMA 293, 18681874.CrossRefGoogle ScholarPubMed
4.Finkelstein, EA, Fiebelkorn, IC & Wang, G (2004) State-level estimates of annual medical expenditures attributable to obesity. Obes Res 12, 1824.CrossRefGoogle ScholarPubMed
5.Quesenberry, CP Jr, Caan, B & Jacobson, A (1998) Obesity, health services use, and health care costs among members of a health maintenance organization. Arch Intern Med 158, 466472.CrossRefGoogle ScholarPubMed
6.Wolf, AM & Colditz, GA (1998) Current estimates of the economic cost of obesity in the United States. Obes Res 6, 97106.CrossRefGoogle ScholarPubMed
7.Lawrence, RC, Helmick, CG, Arnett, FC et al. (1998) Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 41, 778799.3.0.CO;2-V>CrossRefGoogle ScholarPubMed
8.Centers for Disease Control and Prevention (2007) National and state medical expenditures and lost earnings attributable to arthritis and other rheumatic conditions – United States, 2003. MMWR Morb Mortal Wkly Rep 56, 47.Google Scholar
9.Madhok, R, Lewallen, DG, Wallrichs, SL, Ilstrup, DM, Kurland, RL & Melton, LJ 3rd (1993) Trends in the utilisation of primary total hip arthroplasty, 1969 through 1990: a population-based study in Olmsted County, Minnesota. Mayo Clin Proc 68, 1118.CrossRefGoogle Scholar
10.Birrell, F, Johnell, O & Silman, A (1999) Projecting the need for hip replacement over the next three decades: influence of changing demography and threshold for surgery. Ann Rheum Dis 58, 569572.CrossRefGoogle ScholarPubMed
11.Powell, A, Teichtahl, AJ, Wluka, AE & Cicuttini, FM (2005) Obesity: a preventable risk factor for large joint osteoarthritis which may act through biomechanical factors. Br J Sports Med 39, 45.CrossRefGoogle ScholarPubMed
12.Wendelboe, AM, Hegmann, KT, Biggs, JJ, Cox, CM, Portmann, AJ, Gildea, JH, Gren, LH & Lyon, JL (2003) Relationships between body mass indices and surgical replacements of knee and hip joints. Am J Prev Med 25, 290295.CrossRefGoogle ScholarPubMed
13.Chang, RW, Pellisier, JM & Hazen, GB (1996) A cost-effectiveness analysis of total hip arthroplasty for osteoarthritis of the hip. JAMA 275, 858865.CrossRefGoogle ScholarPubMed
14.DeFrances, CJ & Hall, MJ (2004) 2002 National Hospital Discharge Survey. Adv Data 129.Google ScholarPubMed
15.Kozak, LJ, DeFrances, CJ & Hall, MJ (2006) National hospital discharge survey: 2004 annual summary with detailed diagnosis and procedure data. Vital Health Stat 13 1209.Google ScholarPubMed
16.Jibodh, SR, Gurkan, I & Wenz, JF (2004) In-hospital outcome and resource use in hip arthroplasty: influence of body mass. Orthopedics 27, 594601.Google ScholarPubMed
17.Epstein, AM, Read, JL & Hoefer, M (1987) The relation of body weight to length of stay and charges for hospital services for patients undergoing elective surgery: a study of two procedures. Am J Public Health 77, 993997.CrossRefGoogle ScholarPubMed
18.Berry, DJ, Kessler, M & Morrey, BF (1997) Maintaining a hip registry for 25 years. Mayo Clinic experience. Clin Orthop Relat Res 6168.CrossRefGoogle ScholarPubMed
19.Charlson, ME, Pompei, P, Ales, KL & MacKenzie, CR (1987) A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40, 373383.CrossRefGoogle ScholarPubMed
20.Long, KH, Bannon, MP, Zietlow, SP, Helgeson, ER, Harmsen, WS, Smith, CD, Ilstrup, DM, Baerga-Varela, Y & Sarr, MG (2001) A prospective randomized comparison of laparoscopic appendectomy with open appendectomy: clinical and economic analyses. Surgery 129, 390400.CrossRefGoogle ScholarPubMed
21.Fong Soohoo, N, Zingmond, DS, Lieberman, JR & Ko, CY (2006) Optimal timeframe for reporting short-term complication rates after total knee arthroplasty. J Arthroplasty 21, 705711.CrossRefGoogle Scholar
22.Batsis, JA, Naessens, JM, Keegan, MT, Huddleston, PM, Wagie, A & Huddleston, JM (2009) Cost-savings of hip arthroplasty patients on specialty orthopedic surgery units. Am J Orthop 38, E5E11.Google Scholar
23.Zizza, C, Herring, AH, Stevens, J & Popkin, BM (2004) Length of hospital stays among obese individuals. Am J Public Health 94, 15871591.CrossRefGoogle ScholarPubMed
24.Thompson, D, Brown, JB, Nichols, GA, Elmer, PJ & Oster, G (2001) Body mass index and future healthcare costs: a retrospective cohort study. Obes Res 9, 210218.CrossRefGoogle ScholarPubMed
25.Lehman, DE, Capello, WN & Feinberg, JR (1994) Total hip arthroplasty without cement in obese patients. A minimum two-year clinical and radiographic follow-up study. J Bone Joint Surg Am 76, 854862.CrossRefGoogle ScholarPubMed
26.Bowditch, MG & Villar, RN (1999) Do obese patients bleed more? A prospective study of blood loss at total hip replacement. Ann R Coll Surg Engl 81, 198200.Google Scholar
27.Schwann, TA, Habib, RH, Zacharias, A, Parenteau, GL, Riordan, CJ, Durham, SJ & Engoren, M (2001) Effects of body size on operative, intermediate, and long-term outcomes after coronary artery bypass operation. Ann Thorac Surg 71, 521530.CrossRefGoogle ScholarPubMed
28.Finkielman, JD, Gajic, O & Afessa, B (2004) Underweight is independently associated with mortality in post-operative and non-operative patients admitted to the intensive care unit: a retrospective study. BMC Emerg Med 4, 3.CrossRefGoogle ScholarPubMed
29.Ray, DE, Matchett, SC, Baker, K, Wasser, T & Young, MJ (2005) The effect of body mass index on patient outcomes in a medical ICU. Chest 127, 21252131.CrossRefGoogle Scholar
30.Prem, KA, Mensheha, N & McKelvey, JL (1965) Operative treatment of adenocarcinoma of the endometrium in obese women. Am J Obstet Gynecol 92, 1622.CrossRefGoogle ScholarPubMed
31.Moulton, MJ, Creswell, LL, Mackey, ME, Cox, JL & Rosenbloom, M (1996) Obesity is not a risk factor for significant adverse outcomes after cardiac surgery. Circulation 94, II87II92.Google Scholar
32.Jiganti, JJ, Goldstein, WM & Williams, CS (1993) A comparison of the perioperative morbidity in total joint arthroplasty in the obese and nonobese patient. Clin Orthop Relat Res 175179.Google ScholarPubMed
33.Valiyeva, E, Russell, LB, Miller, JE & Safford, MM (2006) Lifestyle-related risk factors and risk of future nursing home admission. Arch Intern Med 166, 985990.CrossRefGoogle ScholarPubMed
34.Legner, VJ, Doerner, D, Reilly, DF & McCormick, WC (2004) Risk factors for nursing home placement following major nonemergent surgery. Am J Med 117, 8286.CrossRefGoogle ScholarPubMed
35.Lubitz, JD & Riley, GF (1993) Trends in Medicare payments in the last year of life. N Engl J Med 328, 10921096.CrossRefGoogle ScholarPubMed
36.Waldo, DR, Sonnefeld, ST, McKusick, DR & Arnett, RH 3rd (1989) Health expenditures by age group, 1977 and 1987. Health Care Financ Rev 10, 111120.Google ScholarPubMed
37.Martineau, P, Filion, KB, Huk, OL, Zukor, DJ, Eisenberg, MJ & Antoniou, J (2005) Primary hip arthroplasty costs are greater in low-volume than in high-volume Canadian hospitals. Clin Orthop Relat Res 152156.CrossRefGoogle ScholarPubMed
38.Melton, LJ 3rd (1996) History of the Rochester Epidemiology Project. Mayo Clin Proc 71, 266274.CrossRefGoogle ScholarPubMed
39.US Census Bureau (2000) 2000 Census of Population and Housing. Public Use Microdata Sample, United States. (accessed September 2008).Google Scholar
40.Rowland, ML (1990) Self-reported weight and height. Am J Clin Nutr 52, 11251133.CrossRefGoogle ScholarPubMed
41.Bozic, KJ, Katz, P, Cisternas, M, Ono, L, Ries, MD & Showstack, J (2005) Hospital resource utilisation for primary and revision total hip arthroplasty. J Bone Joint Surg Am 87, 570576.CrossRefGoogle Scholar
42.Iezzoni, LI, Foley, SM, Daley, J, Hughes, J, Fisher, ES & Heeren, T (1992) Comorbidities, complications, and coding bias. Does the number of diagnosis codes matter in predicting in-hospital mortality? JAMA 267, 21972203.CrossRefGoogle ScholarPubMed
43.Mears, SC, Bawa, M, Pietryak, P, Jones, LC, Rajadhyaksha, AD, Hungerford, DS & Mont, MA (2002) Coding of diagnoses, comorbidities, and complications of total hip arthroplasty. Clin Orthop Relat Res 164170.CrossRefGoogle ScholarPubMed
44.Hogan, P, Dall, T & Nikolov, P (2003) Economic costs of diabetes in the US in 2002. Diabetes Care 26, 917932.Google ScholarPubMed
45.Nichols, GA & Brown, JB (2002) The impact of cardiovascular disease on medical care costs in subjects with and without type 2 diabetes. Diabetes Care 25, 482486.CrossRefGoogle ScholarPubMed
46.Andreyeva, T, Sturm, R & Ringel, JS (2004) Moderate and severe obesity have large differences in health care costs. Obes Res 12, 19361943.CrossRefGoogle ScholarPubMed
47.Marzetti, E & Leeuwenburgh, C (2006) Skeletal muscle apoptosis, sarcopenia and frailty at old age. Exp Gerontol 41, 12341238.CrossRefGoogle ScholarPubMed
48.Gallagher, D, Visser, M, Sepulveda, D, Pierson, RN, Harris, T & Heymsfield, SB (1996) How useful is body mass index for comparison of body fatness across age, sex, and ethnic groups? Am J Epidemiol 143, 228239.CrossRefGoogle ScholarPubMed
49.Smalley, KJ, Knerr, AN, Kendrick, ZV, Colliver, JA & Owen, OE (1990) Reassessment of body mass indices. Am J Clin Nutr 52, 405408.CrossRefGoogle ScholarPubMed
50.Vincent, HK, Weng, JP & Vincent, KR (2007) Effect of obesity on inpatient rehabilitation outcomes after total hip arthroplasty. Obesity (Silver Spring) 15, 522530.CrossRefGoogle ScholarPubMed
Figure 0

Table 1 Characteristics of three analyses examining resource utilization within the indexed surgical episode in patients undergoing total hip arthroplasty

Figure 1

Table 2 Characteristics of 5642 patients undergoing elective hip arthroplasty by BMI category*

Figure 2

Table 3 Univariate unadjusted length of stay and cost centre estimates of 5642 elective total hip arthroplasty patients by BMI category*

Figure 3

Table 4 Multivariate analysis of length of stay and cost centre estimates of 5642 elective total hip arthroplasty patients by BMI category*

You have Access
Cited by

Save article to Kindle

To save this article to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle.

Note you can select to save to either the or variations. ‘’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

Impact of body mass on hospital resource use in total hip arthroplasty
Available formats

Save article to Dropbox

To save this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Dropbox account. Find out more about saving content to Dropbox.

Impact of body mass on hospital resource use in total hip arthroplasty
Available formats

Save article to Google Drive

To save this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Google Drive account. Find out more about saving content to Google Drive.

Impact of body mass on hospital resource use in total hip arthroplasty
Available formats

Reply to: Submit a response

Please enter your response.

Your details

Please enter a valid email address.

Conflicting interests

Do you have any conflicting interests? *