Eur J Health Econ DOI 10.1007/s10198-016-0803-4
ORIGINAL PAPER
Economies of scale: body mass index and costs of cardiac surgery in Ontario, Canada Ana P. Johnson1,3 • Joel L. Parlow2 • Brian Milne2 • Marlo Whitehead3 Jianfeng Xu3 • Susan Rohland3 • Joelle B. Thorpe2
•
Received: 6 November 2015 / Accepted: 27 April 2016 Ó Springer-Verlag Berlin Heidelberg 2016
Abstract An obesity paradox has been described, whereby obese patients have better health outcomes than normal weight patients in certain clinical situations, including cardiac surgery. However, the relationship between body mass index (BMI) and resource utilization and costs in patients undergoing coronary artery bypass graft (CABG) surgery is largely unknown. We examined resource utilization and cost data for 53,224 patients undergoing CABG in Ontario, Canada over a 10-year period between 2002 and 2011. Data for costs during hospital admission and for a 1-year follow-up period were derived from the Institute for Clinical Evaluative Sciences, and analyzed according to pre-defined BMI categories using analysis of variance and multivariate models. BMI independently influenced healthcare costs. Underweight patients had the highest per patient costs ($50,124 ± $36,495), with the next highest costs incurred by morbidly obese ($43,770 ± $31,747) and normal weight patients ($42,564 ± $30,630). Obese and overweight patients had the lowest per patient costs ($40,760 ± $30,664 and $39,960 ± $25,422, respectively). Conversely, at the population level, overweight and obese patients were responsible for the highest total yearly population costs to the healthcare system ($92 million and $50 million,
& Joel L. Parlow
[email protected] 1
Department of Public Health Sciences, Queen’s University, Kingston, ON K7L 3N6, Canada
2
Department of Anesthesiology and Perioperative Medicine, Queen’s University, Kingston General Hospital, 76 Stuart Street, Kingston, ON K7L 2V7, Canada
3
Institute for Clinical Evaluative Sciences Queen’s, Queen’s University, Kingston, ON K7L 3N6, Canada
respectively, compared to $4.2 million for underweight patients). This is most likely due to the high proportion of CABG patients falling into the overweight and obese BMI groups. In the future, preoperative risk stratification and preparation based on BMI may assist in reducing surgical costs, and may inform health policy measures aimed at the management of weight extremes in the population. Keywords Obesity paradox Healthcare costs Cardiac surgery outcomes Body mass index Public health JEL Classification I180
Introduction Obesity has been shown to adversely affect many outcomes after surgery. Perioperative mortality, wound infections, stroke, renal failure, respiratory syndromes, prolonged ventilation, pneumonia, sepsis, arrhythmias, pulmonary embolism, and excessive bleeding have all been identified as potential adverse events associated with high body mass index (BMI) [1–7]. These adverse events frequently necessitate prolonged hospitalization, sometimes involving intensive care unit (ICU) stay. In Canada, underweight (BMI \ 18.5 kg/m2), normal weight (BMI = 18.5–24.9 kg/m2), overweight (BMI = 25–29.9 kg/m2), obese (BMI = 30–34.9 kg/m2), and morbidly obese (BMI [ 34.9 kg/m2) individuals comprise 1.4, 30.4, 42.1, 19.8, and 6.2 % of the male population, and 2.6, 41.7, 31.5, 14.4, and 9.8 % of the female population, respectively [8]. With most of the population falling into the latter three BMI categories, population health research has increasingly focused on the rise in health care costs caused by obesity. However, it has been noted recently that
123
A. P. Johnson et al.
patients with a BMI of less than 20 [9] or less than 19 [10] are at the highest risk of adverse outcomes after cardiac surgery. In fact, an ‘‘underweight paradox’’ has been described, whereby patients with a BMI of less than 18.5 have better lung gas exchange, but a longer ICU stay following cardiac surgery than normal weight patients [11]. The differential health outcome between BMI groups after surgery is a strong indication that patients may incur varying levels of cost to the healthcare system depending on their BMI. Several studies linking individuals’ BMI and their hospital usage over a period of time have shown that higher BMIs are associated with higher healthcare costs relative to normal weight individuals in Sweden, Denmark, Canada, and Australia [12–15]. A different relationship is revealed when the study subject pool is restricted to surgical patients. Several studies have shown no substantial influence of BMI on the cost of surgery, including patient populations undergoing total knee arthroplasty [16], radical cystectomy [17], and percutaneous nephrostolithotomy [18]. However, two studies comparing obese (BMI C 30) and non-obese (BMI \ 30) patients report that obese patients incurred higher costs after undergoing partial nephrectomy [19] or radical prostatectomy [20]. It is important to note, however, that underweight patients were not included in any of these surgical cost studies. In two studies that do include underweight patients, it is individuals at the two weight extremes (underweight and obese) who incurred the highest hospital costs after elective noncardiac surgery [21] and after hip and knee replacement surgery [22]. Thus, the relationship between BMI and healthcare costs may depend on the patient population, in terms of the reason for hospitalization and in terms of what BMI categories are included in the analyses. Although a number of studies have examined perioperative complication rates related to obesity [9, 23], few studies have addressed the relationship between BMI and healthcare utilization in patients undergoing cardiac surgery. Drain et al. [24] explored the relationship between BMI and the cost of ICU and ward stay prospectively in 6100 cardiac surgery patients in England, and reported that BMI influenced ICU and ward length of stay (and by extrapolation healthcare costs). Underweight and morbidly obese patients had longer ICU stays, while overweight and obese patients had shorter ICU stays [24] in relation to normal weight patients. Choi et al. [25] compared a small sample of morbidly obese and non-morbidly obese patients (N = 224) undergoing coronary artery bypass grafting (CABG) over a 10-year period. They found similar mortality rates between BMI groups, but with higher healthcare utilization in the morbidly obese group [25]. No studies to date have directly analyzed costs, but have instead used healthcare utilization as a surrogate for costs incurred by cardiac surgery patients. In this study, we explored the
123
relationship between BMI and economic costs following cardiac surgery in a large sample of patients undergoing coronary artery bypass surgery with long-term follow-up, using administrative databases housed at the Institute for Clinical Evaluative Sciences (ICES) in Ontario, Canada.
Methods The study population was comprised of Ontario residents over 18 years of age who had undergone CABG, with or without open chest aortic valve replacement (AVR), between April 1, 2002 and October 31, 2011 in Ontario, for whom sex, age, height, and weight data were accessible, as identified in the Canadian Institute of Health Information (CIHI) Discharge Abstract Database. The date of first cardiac surgery was considered to be the index date, and eligible patients were followed for 1 year with respect to major outcomes. Patients who had other cardiac procedures, such as percutaneous coronary intervention or other valve procedures performed during the same admission were excluded. Cardiac Care Network of Ontario (CCN) data and the following datasets were linked using unique encoded identifiers and analyzed at ICES: Ontario Health Insurance Plan, CIHI Discharge Abstract Database, National Ambulatory Care Reporting System, Same Day Surgery, and the Registered Persons Database. BMI was calculated as weight (kg)/height (m2), and patients were divided into groups: underweight (BMI \ 20 kg/m2), normal weight (BMI 20.0–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2), obese (BMI 30.0–34.9 kg/m2), and morbidly obese (BMI [ 34.9 kg/m2) [26]. The Elixhauser index, a validated measurement score that includes 30 coexisting conditions, was used to adjust for baseline comorbidities that are predictive of long-term mortality [27, 28]. Higher Elixhauser scores reflect a less healthy state. Resource utilization and costs included those incurred during surgical admission for cardiac surgery and during a 1-year period after hospital discharge. Costs included hospitalizations, ambulatory emergency department visits, same day surgeries, physician services, physician visits, laboratory services, and non-physician costs. Data from primary/specialist physician visits, non-physician visits, and laboratory services were calculated as fees paid (cost per visit), and were obtained from the Claims History Database of the Ontario Health Insurance Plan [29]. Costs associated with short episodes of care (mean duration less than 60 days) such as emergency visits, day procedures, medical day/night care, and ambulatory clinics (including outpatient oncology and dialysis services) were estimated from the National Ambulatory Care Reporting System, by multiplying the assigned Day Procedure Group Resource
Economies of scale: body mass index and costs of cardiac surgery in Ontario, Canada
Intensity Weight for each visit by the provincial average cost per weighted case [29]. Similarly, hospital admission costs were estimated from the CIHI Discharge Abstract Database, using the Resource Intensity Weight (RIW) method, a measure of resource utilization intensity, whereby each hospital inpatient is assigned a RIW value representing an average amount of hospital resources (administration, staff, supplies, technology, equipment) utilized; for example, twice as many resources are utilized by an individual with RIW = 2.0 compared with an individual with RIW = 1.0. Individuals classified with particular RIWs have similar resource utilization patterns fitting into statistically and clinically homogenous groups dependent on clinical/administrative data profiles [29–33]. Costs were derived using a top-down allocation method, in which the Ontario Ministry of Health and Long-Term Care (MOHLTC) pays amounts to providers related to relevant sectors of care, mostly hospitals. According to this top-down methodology, hospital costs are attributed to each major health service sector (e.g., same day surgery) according to the Ontario Cost Distribution Methodology (OCDM), developed and used by the Ministry’s Health Data Branch. In turn, health service sector costs are attributed to individual units of utilization recorded in the equivalent database (e.g., Discharge Abstract Database for acute inpatient care). For physician payments, costs constitute fees paid by the payers (MOHLTC or local health integration networks) for the particular service. Resource utilization and cost data for patients were analyzed using analysis of variance (ANOVA), and categorical variables were analyzed using Pearson chi-square or Fisher’s exact tests. Elixhauser scores were dichotomized into scores of B 2 (healthy) and [ 2 (less healthy), based on the median score found in the total study population [34, 35]. Post hoc pairwise comparisons were done using the Tukey’s HSD test for continuous variables. Due to differences between BMI groups in terms of age, gender, and Elixhauser score, multiple regression analyses were carried out to determine the influence of BMI (independent of age, gender, and comorbidities) on surgical admission costs and healthcare costs 1 year post-discharge. Age and Elixhauser scores centred around the median, gender (coded as 0 = male, 1 = female), and BMI categories using normal weight as the reference category were included in the regression model as covariates. The significance level was set at 0.05. Mean and standard deviation (SD) were used for continuous variables, and categorical variables were expressed as percentages. Costs were reported in 2014 Canadian dollars (http://www.stat can.gc.ca), and analyses were performed from the payer’s (MOHLTC) perspective. Data are presented as individual costs (mean cost per patient within each BMI category) and as population costs (total cost per annum for all individuals
within each BMI category during the course of the study). This study was approved by the institutional review boards at Sunnybrook Health Sciences Centre, Toronto, Canada, and at Queen’s University, Kingston, Canada.
Results Patient population and demographics The study population consisted of 53,224 patients (48,005 CABG only and 5219 combined CABG/AVR). On average, the underweight group was the oldest, and the morbidly obese group was the youngest (Table 1). Additionally, Elixhauser scores were influenced by BMI, with a higher proportion of underweight and morbidly obese patients falling into the less healthy category (scores [ 2, p \ 0.001). Mean individual healthcare costs Overall healthcare costs (sum of costs incurred during surgical admission for CABG and during 1 year post-discharge) were influenced by BMI (ANOVA, p \ 0.001). Post hoc comparisons revealed that underweight patients had the highest individual healthcare costs ($50,124 ± $36,495), while obese and overweight patients incurred the lowest costs ($40,760 ± $30,664 and $39,960 ± $25,422, respectively). Morbidly obese and normal weight patients generated similar costs ($43,770 ± $31,747 and $42,564 ± $30,630, respectively), and both groups incurred higher costs than obese and overweight patients. Costs related to surgical admission Figure 1a shows the mean cost per individual during surgical admission by BMI category for each year during the study period. Averaging data across all years, underweight patients incurred the highest mean costs during surgical admission compared to all other BMI categories (all p \ 0.05, Table 2). Normal weight and morbidly obese patients incurred higher costs than both obese and overweight patients (p \ 0.05). Post-discharge costs per patient Figure 1b shows the mean cost per individual within 1 year of surgery by BMI category for each year during the study period. As with surgical admission costs, underweight patients incurred the highest mean costs of combined postdischarge healthcare compared to all other BMI categories (p \ 0.05) (Table 2). Morbidly obese and normal weight patients had the next highest post-discharge costs,
123
A. P. Johnson et al. Table 1 Baseline characteristics Variable
Value
Underweight (n = 842)
Normal weight (n = 11,466)
Overweight (n = 23,023)
Obese (n = 12,405)
Morbidly obese (n = 5488)
Total (n = 53,224)
p valuea
Age at admission
Mean ± SDb
68 ± 11
67 ± 10
66 ± 10
64 ± 10
62 ± 9
65 ± 10
\0.001
Weight (kg)
Mean ± SD
53.0 ± 7.3
66.8 ± 8.1
79.8 ± 9.3
92.4 ± 10.9
110.0 ± 18.1
82.6 ± 16.9
\0.001
Sex
Female, n (%)
348 (41.3)
2623 (22.9)
4031 (17.5)
2544 (20.5)
1586 (28.9)
11,132 (20.9)
\0.001
Male, n (%)
494 (58.7)
8843 (77.1)
18,992 (82.5)
9861 (79.5)
3902 (71.1)
42,092 (79.1)
Median (IQRc)
3 (1–4)
2 (1–3)
2 (1–3)
2 (1–3)
3 (2–4)
2 (1–3)
\0.001 \0.001
Elixhauser score (continuous)
Elixhauser score B2 n (%) [2
414 (49)
7029 (61)
14,956 (65)
7384 (60)
2509 (46)
32,292 (61)
428 (51)
4437 (39)
8067 (35)
5021 (40)
2979 (54)
20,932 (39)
a
p values were generated using ANOVA for all variables except Elixhauser score (continuous), which was analyzed using Kruskal–Wallis nonparametric tests, and sex and Elixhauser score (dichotomized), which were both analyzed using chi-square tests b c
SD standard deviation IQR interquartile range
Fig. 1 Mean individual healthcare costs in 2014 Canadian dollars a during coronary bypass surgery admission, and b within 1 year of surgery, by BMI category by year from 2002 to 2011
respectively, with overweight and obese patients incurring the lowest costs during the year after discharge. Individual post-discharge healthcare costs are shown in Table 3. For emergency department visits, underweight and morbidly obese patients had the highest costs (p \ 0.05 compared to all other BMI groups), while normal weight patients had higher costs than overweight patients (p \ 0.05) and similar costs to obese patients. Underweight patients had the highest costs for fee-for-service (FFS) physician visits (p \ 0.05), while overweight and obese patients had the lowest. With respect to readmission costs during 1 year post-discharge, underweight patients incurred significantly higher costs than all other BMI groups (all p \ 0.05). Normal weight and morbidly obese patients had
123
similar readmission costs, and both groups had higher costs than overweight and obese patients (all p \ 0.05). Total population costs Although overweight and obese patients had similarly low mean individual costs, and underweight patients had the highest mean individual costs compared to all other BMI groups (Table 2), the vast preponderance of overweight and obese patients in our sample resulted in these two BMI groups incurring a substantially higher total yearly population cost to the health system ($92 million for overweight patients and $50 million for obese patients) than underweight patients ($4.2 million) (Fig. 2).
Economies of scale: body mass index and costs of cardiac surgery in Ontario, Canada Table 2 Mean individual and total population healthcare costs (in 2014 Canadian dollars) during surgical admission and 1 year post-discharge from 2002 to 2011 Variable
Underweight (n = 842)
Normal weight (n = 11,466)
Overweight (n = 23,023)
Obese (n = 12,405)
Morbidly obese (n = 5,488)
p valuea
Mean cost during surgical admissionb
30,526 ± 16,572
28,465 ± 14,352*
27,413 ± 12,560*
27,779 ± 17,844*
28,340 ± 14,533*
\0.001
Mean cost 1 year postdischargeb
19,598 ± 29,494
14,099 ± 23,866*
12,547 ± 19,561*
12,981 ± 21,605*
15,430 ± 25,073* \0.001
Total population cost during surgical admission
25,702,981
326,380,103
631,138,139
344,604,100
155,527,650
N/A
Total population cost 1 year post-discharge
16,501,548
161,655,808
288,863,165
161,032,277
84,678,425
N/A
*Significant difference from underweight BMI within each variable (Tukey HSD, p \ 0.05)
Significant difference from normal weight BMI within each variable (Tukey HSD, p \ 0.05)
a
ANOVA
b
Mean values are ± SD
Table 3 Mean individual healthcare costs, in 2014 Canadian dollars, within 1 year post-surgery Cost Item Emergency department visits
Underweight (n = 842)
Normal weight (n = 11,466)
495 ± 843
392 ± 723*
Overweight (n = 23,023)
Obese (n = 12,405)
Morbidly obese (n = 5,488)
p value
380 ± 702*
455 ± 783
\0.001
359 ± 693*
Same day surgery
344 ± 1351
290 ± 1091
280 ± 1147
273 ± 1273
283 ± 1289
0.457
Dialysis
757 ± 7924
544 ± 6888
300 ± 5099
370 ± 5662
339 ± 5569
0.002
Laboratory billings Physician fee for service billings Family physician capitation Non-physician billings Readmission
236 ± 245
248 ± 262
254 ± 254
\0.001
3075 ± 4382
2612 ± 3578*
2429 ± 2479*
2451 ± 3476*
2615 ± 2885*
\0.001
76 ± 131
85 ± 133
73 ± 120
73 ± 117
74 ± 114
\0.001
41 ± 139 8813 ± 16,005
36 ± 113 6377 ± 16,015*
239 ± 239
240 ± 244
36 ± 108 5761 ± 12,057*
34 ± 99 5844 ± 13,837*
36 ± 117 6997 ± 16,889*
0.312 \0.001
Post hoc pairwise comparisons are shown for comparisons to underweight and normal weight patients for variables of interest All values are mean ± standard deviation *Significant difference from underweight BMI within each variable (Tukey HSD, p \ 0.05)
Significant difference from normal weight BMI within each variable (Tukey HSD, p \ 0.05)
a
All items are calculated after discharge date, including the date of discharge, for 1 year
Multivariate analyses
Discussion
Multiple regression analyses adjusting for potential confounding influences of age, gender, and Elixhauser score were performed to assess the independent influence of BMI on healthcare costs. Analyses confirmed an effect of BMI on total healthcare costs independent of patient age, gender, and comorbidity score (Table 4). Specifically, relative to normal weight patients, underweight patients were substantially more expensive, overweight and obese patients were significantly less expensive, and morbidly obese patients incurred similar costs.
The relationship between BMI and outcome after cardiac surgery has been previously studied [9–11, 23]; however, the impact of BMI on healthcare system costs is largely unknown. The current study confirmed overall differences in costs of care of CABG surgery during surgical admission and 1 year after hospital discharge between BMI groups, independent of patient gender, age, and comorbidities (Elixhauser score). In comparison with the normal weight group, underweight patients experienced the highest mean costs, and overweight and obese patients experienced the
123
A. P. Johnson et al. Fig. 2 Coronary bypass surgery costs, including hospital admission and during 1 year follow-up, by BMI group (Canadian dollars). The inner circle represents the mean cost per patient (in thousands); the outer circle represents the average annual total population cost (in millions). The overweight, obese, and morbidly obese groups incurred 76 % of total healthcare costs
4.2
24.0 48.8 43.8
50.1 Underweight (n=842) Normal Weight (n=11,466)
50.6
Overweight (n=23,023) Obese (n=12,405) Morbidly Obese (n=5,488) 40.8 42.6
40.0 92.0
Table 4 Multiple regression analyses assessing the influence of BMI on mean healthcare cost, relative to normal weight patients, independent of patient age, gender, and Elixhauser score Parameter
Estimate
Standard error
p value
Normal
0.00
–
–
Underweight
4049.57
944.80
\0.01
Overweight
-1193.17
302.67
\0.01
Obese
-1163.47
344.39
\0.01
BMI
Morbidly obese
-848.95
441.69
0.05
Adjusted p for overall BMI
–
–
\0.01
lowest mean costs, while morbidly obese patients did not generate more cost than normal weight patients. Underweight patients undergoing cardiac surgery have elevated rates of morbidity and mortality [9]. In this group, increased comorbidities, higher proportion of females, small coronary vessels, implantation of smaller sized grafts and valves, and increased hemodilution have been proposed as possible causes [36–39]. Non-cardiovascular causes may also be possible contributors, according to epidemiological studies [40, 41]. Based on our previous work [9], underweight patients were the oldest, had a higher frequency of females, and the highest rates of peripheral vascular disease, chronic obstructive pulmonary disease, elevated creatinine, dialysis, cerebrovascular disease, congestive heart failure, and impaired left ventricular function. Underweight patients also experienced the longest ICU and hospital stay, and the highest rates of
123
reoperation, dialysis, myocardial infarction, blood (red cell, platelet) transfusion, and mortality. The morbidly obese group was the youngest, experienced the highest rates of diabetes and hypertension, and had the longest surgery duration. For both of these groups, BMI independently influenced mortality when adjusting for age, gender, type of surgery, congestive heart failure, and cerebrovascular disease [9]. One interpretation of these data is that overweight patients may be better able to cope with operative and postoperative stress as a result of high percentages of body fat, higher serum albumin concentration, and improved nutrition [37]. Morbidity and mortality may be strongly influenced by nutritional status overall [42–45]. Poor nutritional status is not only associated with increased morbidity and mortality in hospital patients, but it is also associated with increased hospital costs. In one study, malnourished patients incurred significantly higher costs of hospitalization than well-nourished patients, which may be a result of an increased length of hospital stay [45]. The correlative nature of this study does not allow us to determine the cause of the healthcare cost discrepancies between BMI groups. However, multiple regression analyses suggest that BMI impacts the cost of healthcare after cardiac surgery independent of potentially confounding factors such as underlying health (Elixhauser score), gender, and age. In a previous study, we confirmed the existence of an obesity paradox, with underweight and morbidly obese cardiac surgery patients having a higher incidence of adverse outcomes and mortality, and obese patients having the lowest incidence of complications and mortality [9]. The increased costs attributed to underweight
Economies of scale: body mass index and costs of cardiac surgery in Ontario, Canada
cardiac surgery patients found here are reflected in the increased hospital resources required by these patients noted previously [9]. Similarly, the reduced incidence of adverse perioperative outcomes in overweight and obese patients found previously [9] may explain the low individual healthcare costs incurred by these patients. Despite several hypotheses outlined above, the reasons for the observed obesity paradox remain unproven, and thus the underlying causes of the differential economic impacts of BMI are speculative. Although obesity rates have been rising across all 34 OECD countries, there is up to a tenfold difference in obesity rates within these countries [46], which could in turn influence which BMI groups incur the greatest total healthcare costs. Nevertheless, an obesity paradox has been observed in several countries, including Canada and the US [9, 47], Spain [23], Scotland [48], and Japan [49]. Therefore, we expect trends in healthcare costs similar to those seen in our Canadian patient population to occur in other countries. A number of previous studies have supported the inclusion of BMI in perioperative risk stratification [4, 37, 50–53]. For example, morbidly obese patients are at higher risk of postoperative infection, potentially due to decreased perfusion of subcutaneous fat tissue [54]. Morbidly obese patients with stable cardiac disease could be referred for preoperative dietetic advice and management. With regard to underweight patients, it remains to be seen whether perioperative nutritional supplementation may reduce morbidity and costs of surgery. With monitoring and optimization of metabolic status prior to, during, and after surgery through pre-loading of carbohydrate and fluid intake and early oral feeding, recovery could be positively affected [55–57]. Similarly, it is possible that health care resources could be more efficiently allocated as a result of preoperative consideration of BMI [24]. Preoperative nutritional support in malnourished surgical patients can significantly reduce the complication rate and length of hospital stay [55], which could in turn lead to reduced healthcare costs incurred by these patients. Changing hospital policies to include an assessment of nutritional status and BMI prior to surgery, and implementing a nutritional program for underweight patients is advisable. For underweight patients, resources directed to optimizing nutritional status prior to surgery may reduce overall costs, although such a strategy must be confirmed in future studies. We must keep in mind, however, that due to the disproportionate number of overweight and obese individuals in the population, these individuals have a much larger impact on the healthcare system in terms of resource use and cost, despite mean individual costs being low compared to normal weight patients. As such, programs to support weight loss
in individuals with high BMI would still have a large economic benefit. In our multivariate analysis, we considered normal weight individuals to be our reference group; however, morbidly obese patients did not generate more costs than normal weight patients. Based on improved outcomes in overweight and obese patients after cardiac surgery compared to normal weight patients, we have previously questioned the traditional definitions of BMI categories as risk factors in this patient population [9]. Our current cost analysis, indicating that overweight and obese patients also incur lower healthcare costs than normal weight patients, again suggests reconsidering the meaning of the traditional definitions of BMI categories in certain populations. The current study has a number of strengths. First, the large sample size allowed robust analysis between different BMI categories, despite the relative paucity of underweight patients (1.6 %) in this population. We also analyzed the effect of BMI independent of a number of confounding variables associated with BMI, such as age, gender and comorbidities. The data available through the ICES databases also allowed the assessment of costs associated with both the surgical admission and a 1-year postoperative follow-up period. There are some limitations to this study, including its retrospective and correlative nature, which precludes us from making definitive statements about the direction of causation. Additionally, since healthcare systems differ across countries, the results reported herein may not reflect healthcare costs incurred by cardiac surgery patients outside of Ontario, Canada. Furthermore, the prevalence of obesity in the population under study impacts the distribution across BMI groups of total versus individual healthcare costs incurred after cardiac surgery.
Conclusions and future research We have demonstrated that, relative to normal weight patients, underweight patients have higher mean costs related to coronary artery surgery while overweight and obese patients have lower costs. This effect occurred independently of age, gender and comorbidity scores. Although overweight and obese patients incur the lowest mean costs compared to other BMI groups, the preponderance of overweight and obese patients in this population leads to the highest overall costs to the health care system. Preoperative risk stratification and preparation with regard to BMI may ultimately assist in reducing overall costs of surgery, as well as informing health policy measures aimed at management of extremes of weight in the population. Future research should focus on understanding the causes underlying the obesity paradox and its economic impacts,
123
A. P. Johnson et al.
as well as developing policies to address the costs incurred disproportionately by underweight patients. Acknowledgments The authors are grateful to Shahriar Khan for his help with statistical analyses. This study was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The opinions, results and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by ICES or the MOHLTC is intended or should be inferred. Parts of this material are based on data and information compiled and provided by the Canadian Institute of Health Information (CIHI) and the Cardiac Care Network of Ontario (CCN). However, the analyses, conclusions, opinions, and statements expressed herein are those of the authors, and not necessarily those of CIHI or CCN. Compliance with ethical standards
10.
11.
12.
13.
14.
Funding sources Internal funding sources were used for this project. Conflict of interest The authors declare that they have no conflict of interest.
15.
References 16. 1. Mokdad, A.H., Serdula, M.K., Dietz, W.H., Bowman, B.A., Marks, J.S., Koplan, J.P.: The spread of the obesity epidemic in the United States, 1991–1998. J. Am. Med. Assoc. 282, 1519–1522 (1999) 2. Mokdad, A.H., Ford, E.S., Bowman, B.A., Dietz, W.H., Vinicor, F., Bales, V.S., Marks, J.S.: Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. J. Am. Med. Assoc. 289, 76–78 (2003) 3. Edwards, F.H., Carey, J.S., Grover, F.L., Bero, J.W., Hartz, R.S.: Impact of gender on coronary bypass operative mortality. Ann. Thorac. Surg. 66, 125–131 (1998) 4. Birkmeyer, N.J., Charlesworth, D.C., Hernandez, F., Leavitt, B.J., Marrin, C.A., Morton, J.R., Olmstead, E.M., O’Connor, G.T.: Northern New England Cardiovascular Disease Study Group: obesity and risk of adverse outcomes associated with coronary artery bypass surgery. Circulation 97, 1689–1690 (1998) 5. Prabhakar, G., Haan, C.K., Peterson, E.D., Coombs, L.P., Cruzzavala, J.L., Murray, G.F.: The risks of moderate and extreme obesity for coronary artery bypass grafting outcomes: a study from the Society of Thoracic Surgeons’ database. Ann. Thorac. Surg. 74, 1125–1130 (2002) 6. Moulton, M.J., Creswell, L.L., Mackey, M.E., Cox, J.L., Rosenbloom, M.: Obesity is not a risk factor for significant adverse outcomes after cardiac surgery. Circulation 94(Suppl 9), II87–II92 (1996) 7. Schwann, T.A., Habib, R.H., Zacharias, A., Parenteau, G.L., Riordan, C.J., Durham, S.J., Engorren, M.: Effects of body size on operative, intermediate, and long-term outcomes after coronary artery bypass operation. Ann. Thorac. Surg. 71, 521–530 (2001) 8. Measured adult body mass index (BMI) by age group and sex, household population aged 18 and over excluding pregnant females, Canada (excluding territories). http://www5.statcan.gc. ca/cansim/a26?lang=eng&id=1050507&p2=46 Statistics Canada (2008) [Accessed March 16, 2016] 9. Johnson, A.P., Parlow, J.L., Whitehead, M., Xu, J., Rohland, S., Milne, B.: Body mass index, outcomes, and mortality following
123
17.
18.
19.
20.
21.
22.
23.
24.
cardiac surgery in Ontario. Canada. J. Am. Heart Assoc. 4, 1–13 (2015) Engel, A.M., McDonough, S., Smith, J.M.: Does an obese body mass index affect hospital outcomes after coronary artery bypass graft surgery. Ann. Thorac. Surg. 88, 1793–1800 (2009) Ranucci, M., Ballota, A., La Rovere, M.T., Castelvecchio, S.: Surgical and Clinical Outcome Research (SCORE) Group: postoperative hypoxia and length of intensive care unit stay after cardiac surgery: the underweight paradox. PLoS ONE 9, 1–8 (2014) ¨ degaard, K., Berglund, G., Nilsson, J.-A., Borg, S., Persson, U., O Nilsson, P.A.: Obesity, survival, and hospital costs—findings from a screening project in Sweden. Value Health 8, 562–571 (2005) Folmann, N.B., Bossen, K.S., Willaing, I., Sorensen, J., Andersen, J.S., Ladelund, S., Jorgensen, T.: Obesity, hospital services use and costs. Adv Health Econ Health Serv Res. 17, 319–332 (2007) Tarride, J.-E., Haq, M., Taylor, V.H., Sharma, A.M., NakhaiPour, H.R., O’Reilly, D., Xie, F., Dolovich, L., Goeree, R.: Health status, hospitalizations, day procedures, and physician costs associated with body mass index (BMI) levels in Ontario. Canada. Clinicoecon Outcomes Res. 4, 21–30 (2012) Korda, R.J., Joshy, G., Paige, E., Butler, J.R.G., Jorm, L.R., Liu, B., Bauman, A.E., Banks, E.: The relationship between body mass index and hospitalisation rates, days in hospital and costs: findings from a large prospective linked data study. PLoS One 10, 1–16 (2015) Batsis, J.A., Naessens, J.M., Keegan, M.T., Huddleston, P.M., Wagie, A.E., Huddleston, J.M.: Body mass index and the impact on hospital resource use in patients undergoing total knee arthroplasty. J. Arthroplasty 25, 1250–1257 (2010) Bagrodia, A., Grover, S., Srivastava, A., Gupta, A., Bolenz, C., Sagalowsky, A.I., Lotan, Y.: Impact of body mass index on clinical and cost outcomes after radical cystectomy. BJU Int. 104, 326–330 (2009) Bagrodia, A., Gupta, A., Raman, J.D., Bensalah, K., Pearle, M.S., Lotan, Y.: Impact of body mass index on cost and clinical outcomes after percutaneous nephrostolithotomy. Urology 72, 756–760 (2008) Bensalah, K., Raman, J.D., Bagrodia, A., Marvin, A., Lotan, Y.: Does obesity impact the costs of partial and radical nephrectomy? J. Urology 179, 1714–1718 (2008) Bolenz, C., Gupta, A., Hotze, T., Ho, R., Cadeddu, J.A., Roehrborn, C.G., Lotan, Y.: The influence of body mass index on the cost of radical prostatectomy for prostate cancer. BJU Int. 106, 1188–1193 (2010) Thomas, E.J., Goldman, L., Mangione, C.M., Marcantonio, E.R., Cook, E.F., Ludwig, L., Sugarbaker, D., Poss, R., Donaldson, M., Lee, T.H.: Body mass index as a correlate of postoperative complications and resource utilization. Am. J. Med. 103, 277–283 (1997) Epstein, A.M., Read, J.L., Hoefer, M.: 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, 993–997 (1987) Lopez-Delgado, J., Esteve, F., Manez, R., Torrado, H., Carrio, M.L., Rodrı´guez-Castro, D., Rarrero, E., Javierre, C., Skaltsa, K., Ventura, J.L.: The influence of body mass index on outcomes in patients undergoing cardiac surgery: does the obesity paradox really exist? PLoS One 10, e0118858 (2015) Drain, A.J., Gerrard, C., Ferguson, J.I., Cafferty, F., Gurprashad, R., Vuylsteke, A.: Does body mass index (BMI) affect cost in cardiac surgery? ‘A pound (£) for pound (lb) analysis’. Interac. Cardiovasc. Thorac. Surg. 5, 282–284 (2006)
Economies of scale: body mass index and costs of cardiac surgery in Ontario, Canada 25. Choi, J.C., Bakaeen, F.G., Cornwell, L.D., Dao, T.K., Coselli, J.S., LeMaire, S.A., Chu, D.: Morbid obesity is associated with increased resource utilization in coronary artery bypass grafting. Ann. Thorac. Surg. 94, 23–28 (2012) 26. Gurm, H.S., Whitlow, P.L., Kip, K.E.: BARI investigators. The impact of body mass index on short- and long-term outcomes inpatients undergoing coronary revascularization investigation (BARI). J. Am. Coll. Cardiol. 39, 834–840 (2002) 27. Elixhauser, A., Steiner, C., Harris, D.R., Coffey, R.M.: Comorbidity measures for use with administrative data. Med. Care 36, 8–27 (1998) 28. Quan, H., Sundararajan, V., Halfon, P., Fong, A., Burnand, B., Luthi, J.C., Saunders, L.D., Beck, C.A., Feasby, T.E., Ghali, W.A.: Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med. Care 43, 1130–1139 (2005) 29. Wodchis, W.P., Bushmeneva, K., Nikitovic, M., McKillop, I.: Guidelines on person-level costing using administrative databases in Ontario. Working Paper Series. Vol 1. Toronto: Health System Performance Research Network; 2013 30. Canadian Coordinating Office for Health Technology Assessment. A guidance document for the costing process, version 1.0. Ottawa: CADTH, 1996 31. Canadian Institute of Health Information. Resource intensity weights. Summary of methodology 1995/96. Ottawa: The Institute; 1995: 1–75 32. Pink, G.H., Bolley, H.B.: Physicians in health care management. 3. Case mix groups and resource intensity weights: an overview for physicians. Can. Med. Assoc. J. 150, 889–894 (1994) 33. Juurlink, D., Preyra, C., Croxford, R., Chong, A., Austin, P., Tu, J., Laupacis, A.: Canadian Institute for Health Information Discharge Abstract Database: a validation study. Institute for Clinical Evaluative Sciences, Toronto (2006) 34. Chandler, P.D., Giovannucci, E.L., Williams, M.A., LeBoff, M.S., Bates, D.W., Hicks, L.S.: Vitamin D deficiency treatment patterns in academic urban medical center. Am. J. Pharm. Benefits 6, e1–e8 (2014) 35. Chen, J.Y., Ma, Q., Everhard, F., Yermilov, I., Tian, H., Mayer, K.M.: HIV screening in commercially insured patients screened or diagnosed with sexually transmitted diseases or blood-borne pathogens. Sex. Transm. Dis. 38, 522–527 (2011) 36. Potapov, E.V., Loebe, M., Anker, S., Stein, J., Bondy, S., Nasseri, B.A., Sodian, R., Hausmann, H., Hetzer, R.: Impact of body mass index on outcome in patients after coronary artery bypass grafting with and without valve surgery. Eur. Heart J. 24, 1933–1941 (2003) 37. Engelman, D.T., Adams, D.H., Byme, J.G., Aranki, S.F., Collins Jr, J.J., Couper, G.S., Allred, E.N., Cohn, L.H., Rizzo, R.J.: Impact of body mass index and albumin on morbidity and mortality after cardiac surgery. J. Thorac. Cardiov. Sur. 118, 866–873 (1999) 38. Loop, F.D., Golding, L.R., MacMillan, J.P., Cosgrove, D.M., Lytle, B.W., Sheldon, W.C.: Coronary artery surgery in women compared with men: analyses of risks and long-term results. J. Am. Coll. Cardiol. 1, 383–390 (1983) 39. O’Connor, G.T., Morton, J.R., Dieh, M.J., Olmestead, E.M., Coffin, L.H., Levy, D.G., Malenka, D.J.: Northern New England Cardiovascular Disease Study Group: differences between men and women in hospital mortality associated with coronary artery bypass graft surgery. Circulation 88, 2104–2110 (1993) 40. Wannamethee, G., Shaper, A.G.: Body weight and mortality in middle-aged British men: impact of smoking. Brit. Med. J. 299, 1497–1502 (1989) 41. Manson, J.E., Stampfer, M.J., Hennekens, C.H., Willett, W.C.: Body weight and longevity: a reassessment. J. Am. Med. Assoc. 257, 353–358 (1987)
42. Anker, S.D., Ponikowski, P., Varney, S., Chua, T.P., Clark, A.L., Webb-Peploe, K.M., Harrington, D., Kox, W.J., Poole-Wilson, P.A., Coats, A.J.S.: Wasting as independent risk factor for mortality in chronic heart failure. Lancet 349, 1050–1053 (1997) 43. Nochioka, K., Sakata, Y., Takahashi, J., Miyata, S., Miura, M., Takada, T., Fukumoto, Y., Shiba, N., Shimokawa, H.: Prognostic impact of nutritional status in asymptomatic patients with cardiac diseases. Circ. J. 77, 2318–2326 (2013) 44. Persson, M.D., Brismar, K.E., Katzarski, K.S., Nordenstro¨m, J., Cederholm, T.E.: Nutritional status using mini nutritional assessment and subjective global assessment predict mortality in geriatric patients. J. Am. Geriatr. Soc. 50, 1996–2002 (2002) 45. Correia, M.I.T.D., Waitzberg, D.L.: The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clin. Nutr. 22, 235–239 (2003) 46. Obesity Update June 2014 OECD: http://www.oecd.org/health/ Obesity-Update-2014.pdf Accessed January 11, 2016 47. Curtis, J.P., Selter, J.G., Wang, Y., Rathore, S.S., Jovin, I.S., Jadbabaie, F., Kosiborod, M., Portnay, E.L., Sokol, S.I., Bader, F., Krumholz, H.M.: The obesity paradox: body mass index and outcomes in patients with heart failure. Arch. Intern. Med. 165, 55–61 (2005) 48. Hastie, C.E., Padmanabhan, S., Slack, R., Pell, A.C.H., Oldroyd, K.G., Flapan, A.D., Jennings, K.P., Irving, J., Eteiba, H., Dominiczak, A.F., Pell, J.P.: Obesity paradox in a cohort of 4880 consecutive patients undergoing percutaneous coronary intervention. Eur. Heart J. 31, 222–226 (2010) 49. Kaneko, H., Yajima, J., Oikawa, Y., Tanaka, S., Fukamachi, D., Suzuki, S., Sagara, K., Otsuka, T., Matsuno, S., Funada, R., Kano, H., Uejima, T., Koike, A., Nagashima, K., Kirigaya, H., Sawada, H., Aizawa, T., Yamashita, T.: Obesity paradox in Japanese patients after percutaneous coronary intervention: an observation cohort study. J. Cardiol. 62, 18–24 (2013) 50. Moulton, M.J., Creswell, L.L., Mackey, M.E., Cox, J.L., Rosenbloom, M.: Reexploration for bleeding is a risk factor for adverse outcomes after cardiac operations. J. Thorac. Cardiovasc. Surg. 111, 1037–1046 (1996) 51. Ridderstolpe, L., Gill, H., Granfeldt, H., Ahlfeldt, H., Rutberg, H.: Superficial and deep sternal wound complications: incidence, risk factors and mortality. Eur. J. Cardiothorac. Surg. 20, 1168–1175 (2001) 52. Lilienfeld, D.E., Vlahov, D., Tenney, J.H., McLaughlin, S.: Obesity and diabetes as risk factors for postoperative wound infections after cardiac surgery. Am. J. Infect. Control 16, 3–6 (1988) 53. Bellchambers, J., Harris, J.M., Cullinan, P., Pepper, J.R.: A prospective study of wound infection in coronary artery surgery. Eur. J. Cardiothorac. Surg. 15, 45–50 (1999) 54. Anaya, D.A., Dellinger, E.P.: The obese surgical patient: a susceptible host for infection. Surg. Infect. 7, 473–480 (2006) 55. Jie, B., Jiang, Z.-M., Nolan, M.T., Zhu, S.-N., Yu, K., Kondrup, J.: Impact of preoperative nutritional support on clinical outcome in abdominal surgical patients at nutritional risk. Nutrition 28, 1022–1027 (2012) 56. Fearon, K.C.H., Luff, R.: The nutritional management of surgical patients: enhanced recovery after surgery. Proc. Nutr. Soc. 62, 807–811 (2003) 57. Weimann, A., Braga, M., Harsanyi, L., Laviano, A., Ljungqvist, O., Soeters, P., Jauch, K.W., Kemen, M., Hiesmayr, J.M., Horbach, T., Kuse, E.R., Vestweber, K.H.: ESPEN guidelines on enteral nutrition: surgery including organ transplantation. Clin. Nutr. 25, 224–244 (2006)
123