Pre-bariatric surgery weight loss requirements and the effect ... - Nature

International Journal of Obesity (2012) 36, 1380–1387 & 2012 Macmillan Publishers Limited All rights reserved 0307-0565/12 www.nature.com/ijo

BARIATRIC AND METABOLIC SURGERY REVIEW

Pre-bariatric surgery weight loss requirements and the effect of preoperative weight loss on postoperative outcome CN Ochner1,2, CL Dambkowski2, BL Yeomans2, J Teixeira3 and F Xavier Pi-Sunyer2 Pre-bariatric surgery requirements vary between surgeons and surgical centers, with standards of practice not yet established. The goal of this systematic review was to summarize and evaluate the available literature on pre-bariatric surgery weight loss requirements and the relation between preoperative weight loss and postoperative outcome. Major databases, including Medline, PubMed and PsychINFO were searched for relevant articles. Case studies, studies420 years old and studies that utilized selfreported body weight data were excluded. Data on the effect of the following was summarized: (1) preoperative requirements on preoperative weight loss; (2) insurance-mandated preoperative requirements; (3) the contingency of receipt of surgery; (4) preoperative weight loss on postoperative weight loss and (5) preoperative weight loss on perioperative and postoperative complication and comorbidity rates. The majority of studies suggest that: (1) current preoperative requirements held by the majority of third party payer organizations in the United States are ineffective in fostering preoperative weight loss; (2) making receipt of surgery contingent upon achieving preoperative weight loss, and meal-replacement diets, may be particularly effective in fostering preoperative weight loss and (3) preoperative weight loss may lead to improvements in at least some relevant postoperative outcomes. However, a preoperative weight loss mandate may lead to the denial of surgery and subsequent health benefits to individuals who are unable to achieve a prespecified amount of weight. Overall, the limited number and quality of prospective studies in this area prohibits the much-needed establishment of standards of practice for pre-bariatric requirements. International Journal of Obesity (2012) 36, 1380–1387; doi:10.1038/ijo.2012.60; published online 17 April 2012 Keywords: gastric bypass; gastric banding; RYGB

INTRODUCTION Bariatric surgery is the only treatment for obesity to demonstrate long-term effectiveness.1–3 Unfortunately, the high cost (ranging from $18 000 to $25 000)4,5 makes self-pay untenable for most bariatric patients, forcing them to rely on third party payment.6 US insurance carriers generally offer coverage for these procedures to candidates meeting the NIH criteria, but often require participation in a 6-month, physician-supervised weight loss regimen.6 However, there is neither a specific diet/ exercise program nor a standard for actual weight loss that are typically suggested or prescribed. Surgical centers may also impose individual preoperative requirements, which vary at the discretion of the surgeon and/or institution. It remains unclear whether prevalent preoperative requirements are effective in producing preoperative weight loss or, furthermore, if preoperative reductions in body weight lead to an improved postoperative outcome. With a growing number of bariatric procedures performed, there is increasing need for consensus on a standard for optimal preoperative practices. The goal of this review was to summarize and evaluate the available literature on pre bariatric surgery weight loss (requirements) and the relation between preoperative weight loss and postoperative outcome.

MATERIALS AND METHODS Data sources and searches We searched Ovid MEDLINE, PubMed and PsychINFO databases to identify relevant articles indexed from 1991 to 2011. The search was conducted between May 1, 2011 and December 31, 2011. Supplemental Table 1 contains the specific search strategy. We obtained two additional articles from the reference lists of relevant studies.

Study selection Articles were reviewed by using the eligibility criteria in Supplemental Table 2. Eligible articles were published in English and provided data on pre-bariatric surgery weight loss, pre-bariatric surgery weight loss requirements, or the relation between pre-bariatric surgery weight loss and peri/post operative outcomes. Studies indexed prior to 1991 were excluded to limit the review to research from the past 20 years, and because bariatric surgery procedures and requirements were not well established prior to the NIH consensus statement in 1991.7 Individual case reports, case series and studies using self-reported body weight data were excluded. Given the relatively small number of studies published in this area, no further exclusion criteria were imposed. Three investigators reviewed citations identified through literature searches. We retrieved the full text of potentially relevant articles for further review. Our systematic search identified 10 830 titles for review (Supplemental Figure 1). Duplicate titles were eliminated from further consideration. Authors independently categorized the articles as being

1 Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, NY, USA; 2New York Obesity Nutrition Research Center, St. Luke’s Roosevelt Hospital, Columbia University College of Physicians and Surgeons, New York, NY, USA and 3Center for Weight Loss Surgery, Department of Minimally Invasive Surgery, St. Luke’s Roosevelt Hospital, Columbia University College of Physicians and Surgeons, New York, NY, USA. Correspondence: Dr CN Ochner, Columbia University College of Physicians and Surgeons, New York Obesity Nutrition Research Center, St. Luke’s Roosevelt Hospital Center, 1111 Amsterdam Ave, Babcock 1020, New York, NY 10025, USA. E-mail: [email protected] Received 16 January 2012; revised 7 March 2012; accepted 11 March 2012; published online 17 April 2012

Pre-bariatric surgery weight loss CN Ochner et al

1381 ineligible, possibly eligible, or probably eligible. Categorization was conducted blind to article authors and affiliations. Articles were initially categorized by title. Titles that all authors agreed were ineligible were eliminated, leaving 162 potentially eligible articles. Abstracts from the remaining articles were then evaluated using the same criteria, leaving 35 potentially eligible articles, which underwent full-text review by all three authors. After discussion and agreement by the authors, 26 articles met criteria for inclusion. The references of these articles were reviewed in the same manner, yielding two additional articles. Finally, one additional article was suggested by a reviewer, for a total of 29 included articles.

Data extraction and quality assessment From each study, we abstracted the following characteristics: study objective, design, methods, participant demographics (sex, age, baseline illness), participant eligibility and exclusion criteria, number of participants, intervention and follow-up duration, outcomes assessed, analytic method used, controlled variables included and results. The quality of each study was rated as ‘good,’ ‘fair’ or ‘poor’ on the basis of US Preventive Services Task Force criteria.8 Good or fair quality evidence includes studies of sufficient design and quality to provide an unbroken chain of evidencesupported linkages, generalizable to the general population, that connect the intervention with health outcomes. Poor evidence contains a formidable break in the evidence chain such that the connection between the intervention and health outcomes is uncertain. As there was a high number of articles rated ‘fair’, we further sub-divided articles rated as ‘fair’ into three ratings; ‘fair/good,’ ‘fair’ and ‘fair/poor’ to increase specificity. When initial individual reviewer quality ratings did not align, consensus was reached through discussion with all authors.

Data synthesis and analysis Heterogeneity within the available literature precluded the ability to conduct a formal quantitative synthesis of article findings. Therefore, we qualitatively synthesized the results, stratified by groupings outlined below, and presented review data as a narrative summary.9

RESULTS Effect of preoperative requirements on preoperative weight loss Preoperative requirements, stemming from the patient’s chosen surgical center and/or insurance carrier, range in structure and may or may not include specified diet prescriptions, physical activity, nutrition counseling or actual weight loss. Below we summarize and discuss studies of preoperative weight loss requirements imposed by insurance carriers10–13 and surgery centers.2,10–20 Few studies have evaluated the effect of current insurancemandated preoperative requirements, irrespective of actual preoperative weight change. The only relevant prospective study10 found no benefit from a X6-month-mandated preoperative medical weight loss program. Kuwada et al.10 reported no difference in preoperative percent excess body weight loss (%EBWL) in patients participating vs not participating in the 6-month preoperative program. We and one other group published relevant retrospective studies, both reporting significant weight gain in patients who successfully completed insurance-mandated preoperative evaluations.11,12 Jamal et al.11 reported a mean weight gain of 2.0±7.5 kg in 72 patients completing 13 weeks of preoperative dietary counseling. Similarly, we reported a body weight increase of 3.7±5.9 kg in 94 patients completing 6 consecutive monthly preoperative weigh-ins.12 These studies suggest that the common preoperative requirements of third party payer organizations are ineffective in fostering preoperative weight loss. Studies of preoperative requirements of medical centers yield more positive results. Some2,16,18 but not all12 studies of preoperative weight loss from surgical centers, which provided ongoing diet/nutritional counseling and clear weight loss prescriptions (typically 10% excess body weight (EBW)), reported overall mean weight losses. Unfortunately, duration of the preoperative intervention was either not specified2,16 or varied & 2012 Macmillan Publishers Limited

between patients,2,18 making it difficult to compare effectiveness across programs. The effect of the duration of the preoperative weight loss period on pre (or post) operative weight loss remains unknown. Although limiting the duration of the preoperative weight loss period could have a positive effect on adherence, it would also lead to less weight loss or require a more restrictive diet. Several studies17,19–21 demonstrate the effectiveness of preoperative all-liquid meal-replacement diets, which provide greater control over caloric and nutritional intake. All-liquid diets tend to deliver fewer calories than conventional diets, reducing the length of time required to meet particular preoperative weight loss goals. The increased dietary structure with meal-replacement diets may also help account for the effectiveness of these programs, as compliance and weight loss typically increase concurrently with amount of structure provided.22 However, the negative finding (no net preoperative weight loss) reported by Liu et al.13 was from a highly structured preoperative weight loss program tailored to each individual by a multi-disciplinary team, in addition to patients being given the explicit goal of losing 10% EBW prior to surgery.13 A summary of relevant articles and findings is provided in Table 1. Contingency of receipt of surgery on weight loss An additional factor infrequently described in the literature, which may impact the effectiveness of preoperative weight loss programs, is whether surgery is contingent upon meeting a prespecified amount of weight loss. Although the duration of preoperative weight loss periods varied between patients,17 or was not specified,16 studies where surgery was reportedly withheld if patients did not lose weight reported higher mean preoperative weight loss (7%15 and 15%16 total body weight) relative to other comparable programs.2,13,18 In the only viable randomized control trial,15 the authors state that surgery was contingent upon achieving 10% TBWL (total body weight loss) prior to surgery. However, the study reports postoperative data on patients who failed to achieve this requirement. Patients in this trial were led to believe that surgery was dependent on achieving the prespecified amount of weight loss, but surgery was not actually withheld at the end of the preoperative intervention period. Although the length of the preoperative weight loss period was unspecified, it is important to note that the preoperative weight loss observed in this study (8% total body weight (TBW)) was comparable to that seen in studies where surgery was contingent upon preoperative weight loss,16,17 and greater than in studies where a particular amount of preoperative weight loss was not specified or was suggested only as a goal.11–13,23 However, the question of which preoperative requirements are most effective in fostering preoperative weight loss is important only if preoperative weight loss is related to postoperative outcomes, including weight loss, medical comorbidities and complication rates. Relation between pre- and postoperative changes in body weight Findings from studies of the relation between pre- and postoperative changes in body weight range from a positive relationship (preoperative weight loss associated with greater postoperative weight loss) to a negative relationship (preoperative weight loss associated with less postoperative weight loss) and many in between (no relationship). Further complicating the issue, several studies report a relationship in certain circumstances but not others (for example, a relationship at 3-month, but not 6-month, follow-up). Several studies15,16,18,24,25 indicate a positive effect of preoperative weight loss on postoperative weight loss. Alvarado et al.16 retrospectively examined whether preoperative weight loss was related to postoperative reductions in EBWL at 1-year International Journal of Obesity (2012) 1380 – 1387


1382 Table 1.

Pre-bariatric surgery requirements and preoperative weight changea

Trial/study

Design

Alami et al.15

RCT

Van Nieuwenhove et al.21 Alger-Mayer et al.18

RCT

Colles et al.19

Prospective

Fris et al.

20

Prospective

Prospective

Kuwada et al.10

Prospective

Martin et al.14

Prospective

Alvarado et al.16

Retrospective

11

Sample sizeb

Preoperative requirements or recommendationsc

61 (10:51) 274 (81:193) 150 (30:120) 32 (19:13) 40 (4:36) 440 (61:379)

10% TBWL (variable; mean ¼ 5±3 months) 800 kcal per day liquid diet (14 days) Goal of 10% TBWL (unspecified) 456 kcal per day liquid diet (12 weeks) 456 kcal per day liquid diet (14 days) MMP vs control in addition to 1300 kcal per day liquid diet. (6 months)

100 (16:84) 90 (9:81) 324 (73:251)

420 kcal per day liquid diet (4–16 weeks) Instructed to lose 10% EBW (unspecified) Insurance-mandated dietary counseling program (13 weeks) Comprehensive diet and exercise program (variable; range unspecified) Insurance-mandated 6 consecutive monthly weigh-ins (6 months) Unspecified liquid diet (7–24 weeks) Multidisciplinary program (0.2–12 months)

Jamal et al.

Retrospective

Liu et al.13

Retrospective

95 (19:76)

Ochner et al.12

Retrospective

100 (17:83)

Pekkarinen and Mustajoki17 Still et al.2

Retrospective

28 (12:16) 884 (192:692)

Retrospective

Preoperative weight change mean (±s.d.) 8%; range ¼ 1 to 16% TBW 4.9±3.6 kg

P-value

Quality

—

Fair/poor

o 0.001

Good

10±7% TBW at 3 years; 10±7% TBW at 4 years 11%; range ¼ 1 to 19% TBW 4±1% TBW

—

Fair/good

o 0.001

Fair/good

o 0.001

Fair/good

RYGB patients MMP: 7.7% EBW Non-MMP: 6.6% EBW Banding patients MMP: 7.5% EBW Non-MMP: 5.7% EBW 19±11% EBW

—

Fair/good

0.001

Poor

7%; range ¼ 0 to 24% TBW 2.0±7.5 kg

—

Fair

—

Fair/good

50% lost weight, 5% EBW; 50% gained weight, þ 5% EBW þ 3.7±5.9 kg; þ 3±4% TBW

—

Fair

o 0.001

Fair

15% TBW

—

Fair/good

83% lost weight; 17% gained weight

—

Fair/poor

Abbreviations: EBW, excess body weight; MMP, mandated medical program; —, not reported; RCT, randomized control trial; TBW, total body weight; TBWL, total body weight loss. aArticles reported in order of rigor of methodological design: randomized control trial, followed by prospective study, followed by retrospective study. bSample size for pre-op weight change analyses reported. Total and (M:F) reported. cDescription and (duration) reported.

follow-up in 90 Roux-en-Y gastric bypass (RYGB) patients. The study reported that preoperative TBWL positively predicted postoperative EBWL, such that preoperative weight loss of 1% of initial weight was associated with an increase of 2% in postoperative EBWL (Po0.05). Alger-Mayer et al.18 reported positive correlations between preoperative TBWL and sustained TBWL at 3-year (Po0.001) and 4-year (P ¼ 0.015) postoperative follow-up, as well as positive correlations between preoperative EBWL and sustained EBWL at 3-year (P ¼ 0.006) and 4-year (P ¼ 0.046) follow-up, in RYGB patients. Ali et al.25 also reported greater postoperative EBWL and reduction in body mass index (BMI) in patients losing vs gaining weight, preoperatively. In the only viable randomized control trial, Alami et al.15 randomly assigned 100 patients seeking RYGB to either a group with a 10% preoperative weight loss requirement or a group with no preoperative weight loss requirement. These authors reported that the preoperative weight loss group lost 44% and 54% EBW at 3 and 6 months post-surgery (respectively), whereas the nonweight loss group lost 33% and 51% at 3 and 6 months postsurgery (respectively). Only the difference at 3 months post was significant; however, insufficient power to detect an effect at 6 months post was noted by the authors, with data available on only 37% of the sample.15 A reanalysis at 1-year follow-up by Solomon et al.24 revealed that patients who lost 45% EBW preoperatively showed lower body weight, BMI, EBWL and percent change in BMI (all P’so0.01) as compared with patients who showed o5% preoperative EBWL. A ‘pilot randomized trial’ was recently published online.26 However, the authors report nil effects with International Journal of Obesity (2012) 1380 – 1387

insufficient power (N ¼ 55 with complete data on only 23 participants) and limited treatment fidelity (participants assigned to the preoperative weight loss program group attended, on average, only two of six scheduled preoperative sessions).26 Thus, conclusions cannot be drawn from this report. Finally, one additional randomized trial was conducted by Van Nieuwenhove et al.,21 but included only a 30-day follow-up. This trial reported no difference in postoperative weight loss between RYGB patients losing weight (4.9±3.6 kg) preoperatively on an 800 kcal per day all-liquid meal diet vs control. However, 30 days post surgery is an insufficient amount of time upon which to base assertions about the effect of preoperative weight loss on postoperative outcome, as the first month post RYGB is typically spent consuming liquids and progressively readjusting to solid foods. Several additional studies report mixed findings.2,27,28 For example, Mrad et al.28 reported a significant relationship between pre- and postoperative weight loss in men (P ¼ 0.02) but not in women. In an analysis of data on 884 RYGB patients grouped by preoperative % EBW change (45 gain, 0–5 gain, 0–5 loss, 5–10 loss, 410 loss), Still et al.2 reported that patients who showed 410% preoperative EBWL were 2.1 times more likely to achieve postoperative EBWL of 70% (P ¼ 0.001) as compared with patients who showed o5% EBWL preoperatively. However, no difference was found between patients who gained or lost 0–10% EBW, and the data presented suggest that no linear relationship was present. Harnisch et al.27 found no difference in EBWL from surgery to 12-month follow-up between RYGB patients who lost 44.5 kg (10 lbs) vs patients who gained weight & 2012 Macmillan Publishers Limited


1383 preoperatively. However, when examining weight change from baseline (beginning of preoperative weight loss intervention) to 12-month follow-up, preoperative weight loss did confer a postoperative weight loss advantage (Po0.001). Thus, weight loss trajectory from surgery to 12-month follow-up was the same irrespective of preoperative weight gain or loss, leading to greater total weight loss (from baseline to 12-month follow-up) in individuals who lost weight prior to surgery. This effect did not persist at 24-month follow-up; however, power may have been limited in 24-month analyses due to attrition.27 There are also a number of studies reporting no relationship between pre- and postoperative weight change,11,13,14,26,27–34 including a prospective analysis14 reporting no difference in weight loss outcome between preoperative dieters (n ¼ 47; losing an average of 7.8 kg) and non-dieters (n ¼ 53; no significant weight loss) across a 4-year follow-up. Several of the studies reporting no relationship do appear to be underpowered but do not provide adequate detail to conduct post-hoc power analyses. One recent retrospective analysis36 reported that individuals losing less weight (o vs 44.5 kg) showed greater postoperative %EBWL. However, these individuals weighed significantly less (P ¼ 0.001) at the time of surgery, and operative weight was not entered as a covariate. No between-group difference in net postoperative weight loss was found.36 Finally, two recent retrospective analyses11,37 found an inverse relationship between pre- and postoperative weight loss, suggesting that patients who lose more weight preoperatively may actually lose less weight following bariatric surgery. Thus, the available study findings range full-spectrum, with the majority split between a positive relationship and no significant relationship. A summary of the relevant articles and findings is presented in Tables 2 and 3. Methodological considerations for articles reporting on the relation between pre- and postoperative weight loss are reported in Table 4. Effect of insurance-mandated preoperative requirements on postoperative outcome Third party payer organizations maintain that the goal of their preoperative requirements is to ‘improve surgical outcomes.’ However, weight loss itself is typically not required.12 The three relevant studies discussed earlier10–13 are the only studies to examine the effect of the presence vs absence of preoperative requirements themselves (irrespective of actual preoperative weight loss) on postoperative weight loss outcome. Our study12 of insurance-mandated preoperative requirements in gastric

Table 2.

banding and bypass patients reported that weight loss at 3 months post surgery did not differ between patients required (n ¼ 94) and not required (n ¼ 59) to complete a 6-month physician-supervised weight loss regimen. Similarly, Kuwada et al.10 tested postoperative weight loss in gastric banding and bypass patients mandated (n ¼ 116) vs not mandated (n ¼ 324) to undergo a 6-month preoperative medical weight loss program. No difference in %EBWL was seen at 6 or 12 months post-surgery. Jamal et al.11 examined gastric banding and bypass patients who were required by their insurance carrier to participate in a preoperative dietary counseling program (n ¼ 72) relative to those with no requirement to participate in a preoperative dietary program (n ¼ 252). Patients did not differ in %EBW, BMI (kg m 2) or body weight at baseline. At 1-year follow-up, the patients with no requirement vs those required to participate in a preoperative dietary program showed a greater %EBWL (Po0.001), lower BMI (Po0.015) and lower body weight (Po0.01). These findings suggest that a preoperative requirement, in and of itself, is unlikely to improve postoperative weight loss, and may even be associated with less-positive postoperative outcomes. Relation between pre- and postoperative changes in other postoperative outcomes Although it is the focus of this review, weight loss itself is not the only important outcome of bariatric surgery. Studies show that preoperative weight loss is associated with a significant reduction in liver size,19,20 which has been related to reductions in surgical time38 and complications.20 In the retrospective examination of 95 RYGB patients, Liu et al.13 reported less intra-operative blood loss (P ¼ 0.03), likelihood of enlarged liver (P ¼ 0.02) and surgical complications (P ¼ 0.02) in patients losing vs gaining weight preoperatively. Benotti et al.39 in re-analyses of the data reported on by Still et al.2 reported a significant inverse relationship between preoperative weight loss and the odds ratio of any surgical complication (P ¼ 0.008). When compared with those patients who showed X10% preoperative EBWL, patients who gained X5% EBW had a 2-fold increased likelihood of complications.39 As with the investigations of the relation between pre- and postoperative weight loss, studies of the relation between preoperative weight loss and other postoperative outcomes report mixed findings. In the retrospective analyses of 90 RYGB patients by Alvarado et al.,16 preoperative weight loss of 45% was associated with shorter operative time (36 min) but not

Relation between pre- and postoperative changes in body weight examined continuously

Study

Design

Sample sizea

Primary analysis

Observed relationship

P-value

Quality

Alger-Mayer et al.18

Prospective correlational

150 (3 years post)

Regression; relation between pre-and post-op change in both TBW and EBW

Positive

Fair/ good

Alvarado et al.16 Carlin et al.35 Mrad et al.28

Retrospective

79 (1 year post) 295 (1 year post) 135 (3 months post)

Positive

Fair

No relationship

NS

Poor Poor

Retrospective

Positive in men; No relationship in women Inverse

0.02 in men; NS in women

Ochner et al.12 Taylor et al.29

Regression; relation between pre-and post-op change in EBW Correlation; relation between pre-op change in % TBW and post-op % EBWL Correlation; relation between pre-and post-op change in TBW (unclear; see Table 4) Regression; relation between pre-and post-op change in TBW Regression; relation between pre-and post-op % change in BMI

0.015 for TBW;o0.001 for EBW o 0.05

0.001

Fair

No relationship

NS

Fair

Retrospective Retrospective

Retrospective

87 (3 months post) 76 (follow-up time varied)

Abbreviations: BMI, body mass index; EBW, excess body weight; EBWL, excess body weight loss; NS, not significant; pre-op, preoperative; post-op, postoperative; TBW, total body weight. aSample size for analyses of relationship between pre-op and first post-op assessment (follow-up time) reported

& 2012 Macmillan Publishers Limited

International Journal of Obesity (2012) 1380 – 1387


1384 Table 3.

Between group effects of pre-op requirements/weight loss on post-op weight loss

Trial/ studya

Design

Sample sizeb

Between group comparison

Postoperative weight loss finding

P-value

Quality

Alami et al.15

RCT

50 (3 months post)

10% weight loss requirement vs no requirement

Pre-op weight loss group lost more EBW at 3 months post-op; No diff at 6 months post-op

Fair/ poor

Solomon et al.24 Parikh et al.26

Reanalysis of RCT data Pilot RCT

44 (1 year post) 23 (6 months post)

Patients who lost weight pre-op lost more weight at 1 year post-op No difference

Martin et al.14

Prospective w/retrospective grouping Retrospective

100 (3 months post)

o 5% pre-op weight loss vsX5% pre-op weight loss Pre-op medically supervised weight management program vs no pre-op program 47.5 kg pre-op weight loss vs no pre-op weight loss

0.029 at 3 months; NS at 6 months 0.009

302 (6 months post)

Pre-op weight loss vs pre-op weight gain

Ali et al.25 Harnisch et al.27

Retrospective

203 (1 year post)

Pre-op weight loss vs pre-op weight gain

Jamal et al.11

Retrospective

253 (1 year post)

Ochner et al.12

Retrospective

87 (3 months post)

Riess et al.36

Retrospective

262 (1 year post)

Taylor et al.29 Kuwada et al.10

Retrospective

76 (varied) 440 (6 months post)

Insurance mandated pre-op requirements vs no pre-op requirements Insurance mandated pre-op requirements vs no pre-op requirements Mandated44.5 kg pre-op weight loss vs 0–4.5 kg pre-op weight loss o 2.5% pre-op weight gain vs 42.5% pre-op weight gain Insurance mandated pre-op Medical program vs nonmandated medical program

Prospective

NS

Fair/ poor Poor

No difference at any post-op time pointc

NS

Poor

Patients who lost weight pre-op showed greater EBWL at 6 and 12 months post-opc Weight loss group had greater % EBWL from program-entry to 12 and 24 months post-opc No requirement group had greater % EBWL at 1 year post-op

o 0.05

Poor

o 0.05

Fair

o 0.001

Fair/ good

No difference at 3 months post-op

NS

Fair

Patients losing less weight pre-op had greater % EBWL at 1 year post-opc No difference

0.001

Poor

NS

Fair

No difference

NS

Fair/ good

Abbreviations: EBW, excess body weight; EBWL, excess body weight loss; NS, not significant; pre-op, preoperative; post-op, postoperative; RCT, randomized control trial. Juńior et al.37 not included, as these authors grouped patients according to postoperative %EBWL (above or below the mean at six different time points). This study reported that individuals above the mean %EBWL at 6, 18, 24 and 36 months (but not at 12 or 48 months) lost less weight during the preoperative period, which was of unspecified length. Van Nieuwenhove et al.21 also not included, as this trial contained only a 30-day follow-up period, which is insufficient to judge the relation between pre- and postoperative weight loss. aArticles reported in order of rigor of methodological design: randomized control trial, followed by prospective study, followed by retrospective study. bSample size for between group analyses at first post-op assessment (follow-up time) reported cLacking control for pre-op between group differences (see Table 4)

associated with postoperative complication rate or correction of co-morbidities. In the analysis by Still et al.,2 individuals showing 45% preoperative EBWL were less likely to have a length of stay 44 days as compared with individuals gaining weight or losing o5% EBW preoperatively (P ¼ 0.03). In the randomized control trial conducted by Alami et al.,15 no differences between the preoperative weight loss and no weight loss groups were seen in complication rate, conversion rate or medical comorbidities at 3 or 6 months post surgery. Harnisch et al.27 reported no difference between patients who lost vs gained weight preoperatively in perioperative complication or conversion rates, or in the remission of diabetes, hypertension or continuous positive airway discontinuation. Patients losing weight did, however, have a significantly shorter operation time (P ¼ 0.02).27 Thus, although there is some evidence that preoperative weight loss may lead to improvements in at least some postoperative outcomes, the evidence is inconsistent. DISCUSSION With the rise in popularity of bariatric procedures for the treatment of obesity, there is an increasing need for the accurate assessment of the effects of preoperative weight loss and the development of standards of practice for preoperative requirements. For such policy development, the overarching International Journal of Obesity (2012) 1380 – 1387

question is whether receipt of bariatric surgery should be contingent upon any prespecified amount of preoperative weight loss. Unfortunately, very few studies15–17 address the vital contingency component. Further, variation in the results and reporting of data in this literature make it difficult to determine what effect preoperative weight loss has on postoperative weight loss, medical comorbidities and (peri and postoperative) complication rates. Despite a large amount of variation, the current literature does illustrate three main points. First, the insurance-mandated preoperative requirements confer no appreciable benefit to bariatric patients. The ultimate goal of such requirements is questionable, particularly considering that no actual weight loss is required. In fact, patients carrying these requirements typically show a significant weight gain in the 6 months prior to surgery.11,12 Aside from creating a challenge, which a certain percentage of otherwise eligible candidates for bariatric surgery will not meet (thus reducing payout costs), the utility of such requirements is not apparent to these authors. Further, the argument could be raised that the responsibility for ensuring that surgical candidates are unable to achieve acceptable weight loss through conventional treatment methods (that is, dieting), and display the ability to follow postoperative diet prescriptions, lies with the health care team as opposed to the insurance company. Second, preoperative weight loss regimens that make receipt of & 2012 Macmillan Publishers Limited


1385 Table 4.

Methodological considerations for articles reporting on relation between pre- and postoperative weight loss

Trial/studya

Methodological considerationsb

Alami et al.15

39% of randomized patients did not undergo surgery Intent-to-treat principle violated; preoperative weight change analyses included only the 61% of patients who underwent surgery Limited statistical power; completer analyses for between group comparisons of EBWL at 3 and 6 months post-op included only 50 and 37 patients (respectively) Surgery made dependent on 10% TBWL, but average weight loss of only 8.2% TBWL reported in patients who underwent surgery; no explanation for discrepancy Only banding patients included Nil results published with insufficient power; total N ¼ 55 Calculation of sample size needed for statistical significance based on two groups (n ¼ 30, n ¼ 30), but patients randomized to three groups (n ¼ 15, n ¼ 15, n ¼ 30) Complete data for less than 50% of patients (n ¼ 23) Intent-to-treat principle violated Limited ecological validity; surgery was not contingent upon completion of preoperative requirements as in real world and preoperative requirements differed greatly from standard insurance-mandated requirements Poor compliance; patients in the mandate group attended an average of only two of six sessions Effect of physical activity not correctly calculated Numerical errors in body and tables Preoperative TBWL defined as the difference between maximum preoperative weight and weight at time of surgery Completer analyses Correlational; no comparison group Age associated with TBWL and EBWL 3 years post-op but was not controlled for in primary analyses examining the relationships between pre- and post-op EBWL and TBWL Completer analyses Mandated patients had requirements over and above typical insurance requirements No data on compliance with requirements; patients who participatedo6 months were eliminated but no mention of how many mandated sessions patients attended Patients grouped retrospectively into successful (47.5 kg) and unsuccessful (o7.5 kg) dieters; unclear why 7.5 kg cutoff chosen Completer analyses Duration of diet varied fromo1 (unspecified) to 4 months Insufficient power to detect between group differences in several outcomes (medical comorbidities) No statistical control for baseline differences between groups; groups (2) differed by sex, TBW, EBW, BMI and psychopathology (all P-valueso0.05) Retrospective Completer analyses No statistical control for baseline between group differences; groups (4) differed by sex, BMI, EBW and TBW (all P-valueso0.05). Without control for baseline differences in body weight/BMI, cannot determine if nil findings are due to there truly being no effect or a failure to control for baseline differences Retrospective Completer analyses Retrospective Correlational; no comparison group Preoperative weight loss (causally) determined by BMI group (o50, 50–60, 460) and post-op EBWL depended upon BMI group (57%, 60%, 68% respectively; Po0.001). Therefore, by design, study was set up to find an inverse relationship; control for BMI group insufficient to accurately evaluate relationship between pre- and post-op weight change Retrospective Completer analyses Possible selection bias; included only patients who deviated from their baseline (initial preoperative evaluation period) weight byX10 lbs (chosen arbitrarily); heavier patients may have been more encouraged to lose weight or encouraged to lose more weight Significant between group differences in baseline body weight likely if no between group differences at time of surgery (after 10 lbs loss vs gain), but not reported or controlled for Group patients according toX10 lb weight loss or gain but graphically report group differences of±10% EBWL just prior to surgery; unclear if analyses performed using absolute weight (lbs) or % EBW Preoperative intervention period varied between patients (no between group difference) Preoperative weight taken 2 weeks prior to surgery Linear relationship between pre- and postoperative weight change not tested Retrospective Completer analyses Post hoc grouping according to patients who were above vs below the mean % EBWL Duration of preoperative period not reported or controlled for No reporting of baseline values or statistical control for potential between group differences in baseline values of any variable Repeated individual comparisons performed without control for artificial alpha inflation. Twenty-six outcome variables examined over six time points ¼ 156 individual tests Retrospective Completer analyses Unclear which analyses performed or reported; stated primary outcome was ‘the correlation of preoperative weight change with post-op weight loss’ (indicating regression), however, analyses performed using ‘a mixed procedure for repeated measures’; later stated that ‘preoperative weight change was empirically defined as42% change from baseline body weight’ but no mention of between group analyses or results, or how42% weight change chosen Interaction with sex not explicitly tested but only sex-stratified results presented

Parikh et al.26

Alger-Mayer et al.18

Kuwada et al.10

Martin et al.14

Ali et al.25

Alvarado et al.16 Carlin et al.35

Harnisch et al.27

Juńior et al.37

Mrad et al.28




1386 Table 4 (Continued ) Trial/studya

Ochner et al.12

Riess et al.36

Taylor et al.29

Methodological considerationsb Regression data not presented; only (single) P-value presented for significant result in men and unclear which test used No statistical control for baseline body weight, irrespective of test used Data from patients undergoing four different types of bariatric procedures combined; no control for procedure type Duration of preoperative evaluation ranged from 100–737 days; no control Retrospective Completer analyses Both groups gained weight preoperatively; limited relevance to the relation between pre and postoperative weight loss Restricted range of preoperative weight change Retrospective Completer analyses Potential sample bias; patients selectively required to lose weight (all patients in weight-loss group) depending on habitus No statistical control for baseline between group differences; groups (2) differed by sex, TBW and BMI (all P-valueso0.05). Without control for baseline differences in body weight/BMI, cannot determine if nil findings are due to there truly being no effect or a failure to control for baseline differences Retrospective Only 17 patients in one of the two groups (gained42.5% initial BMI), which may have limited power Time between surgery and follow-up assessments varied between 9 and 15 months for each patient (no between group difference) Vertical-banded gastroplasty patients only; less relevant to current techniques

Abbreviations: BMI, body mass index; EBW, excess body weight; EBWL, excess body weight loss; N, sample size; n, cell size; TBW, total body weight; TBWL, total body weight loss. Van Nieuwenhove et al.21 not included, as this trial contained only a 30-day follow-up period, which is insufficient to judge the relation between pre- and postoperative weight loss. aArticles reported in order of rigor of methodological design: randomized control trial, followed by prospective study, followed by retrospective study. bNot all points presented in this table should necessarily be considered weaknesses of the included studies, but points that may be important to take into consideration when attempting to derive information regarding particular hypotheses from these studies.

surgery contingent upon preoperative weight loss appear to result in more preoperative weight loss than programs where receipt of surgery is not contingent upon achieving a specified amount of preoperative weight loss. Preoperative weight loss regimens that utilize meal replacements may also be more successful than lessstructured regular food programs. Third, the majority of studies indicate that preoperative weight loss is not likely to be detrimental to bariatric patients, and may lead to improvements in at least some postoperative outcomes (that is, weight loss, complication rate, or medical comorbidities). It is well known that even modest weight loss (irrespective of receipt of bariatric surgery) reduces medical comorbidities.40 Few studies have reported on preoperative changes in medical comorbidities, specifically in the context of a preoperative weight loss regimen. Van de Weijgert et al.41 reported significant reductions in respiratory complaints, impaired glucose tolerance and angina pectoris after preoperative weight loss prior to surgery. Similarly, Aberle et al.38 showed improvements in liver function following preoperative weight loss with Sibutramine. Particularly under the supervision of a physician, nonsurgical weight loss has been shown to be safe.14,17,19–22,42 Therefore, it could be argued that preoperative weight loss poses little risk to patients, and is likely to improve their health preoperatively, regardless of whether it leads to improved postoperative outcome. In an earlier review of this literature,43 the authors suggest that preoperative weight loss should be a necessary component of the preoperative process. However, the review included only a proportion of the available literature, primarily the studies reporting improvements in postoperative outcomes related to preoperative weight loss. It may be true that preoperative weight loss itself may pose little risk, and likely even confer some postoperative benefit, to bariatric patients. However, a preoperative weight loss mandate may still be harmful to some individuals, and careful consideration must be given to all factors before a preoperative weight loss mandate can be recommended. A significant number of obese bariatric candidates would be unable to meet a prespecified amount of preoperative weight loss, thus rendering them ineligible to receive a surgical procedure International Journal of Obesity (2012) 1380 – 1387

that would have likely improved their health and quality of life. Thus, this review may begin to answer the question of whether preoperative weight loss is likely to be beneficial; however, the question remains as to whether the postoperative benefits of preoperative weight loss in individuals able to meet such a requirement outweigh the potential health risk incurred in denying surgery to those who could not. Given the inconsistency and questionable validity of the extant research discussed above on the question of the effect of preoperative weight loss on peri and postoperative outcomes (see Tables 2–4), it is the opinion of these authors that insufficient evidence is currently available to justify a pre-bariatric surgery weight loss mandate. A recommendation for such a mandate from medical associations of influence could potentially result in policy adoption by third party payer organizations, on which the majority of individuals seeking bariatric surgery are dependent. Thus, particularly without compelling evidence for significant positive effects due to preoperative weight loss, it may be unethical to create a policy that will exclude otherwise eligible candidates from a beneficial surgical procedure in the hopes of improving the postoperative outcome of others. With the dramatic increase in the popularity of bariatric procedures in the United States and other nations, there is an understandable sense of urgency to develop a standard of practice for pre-bariatric requirements. However, it is imperative that influential organizations refrain from supporting these potentially harmful conventions until additional data can be obtained, the need for which is clearly illustrated in this review. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS Dr Ochner is supported by a grant from the National Institutes of Health (KL2RR024157). The sponsor had no role in the design and conduct of the review; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.



1387 DISCLAIMER The opinions expressed are the authors’ own. They do not represent any position or policy of the National Institutes of Health, Public Health Service, Department of Health and Human Services or the American Medical Association.

REFERENCES 1 Buchwald H, Williams SE. Bariatric surgery worldwide 2003. Obes Surg 2004; 14: 1157–1164. 2 Still CD, Benotti P, Wood GC, Gerhard GS, Petrick A, Reed M et al. Outcomes of preoperative weight loss in high-risk patients undergoing gastric bypass surgery. Arch Surg 2007; 142: 994–999. 3 Sarwer DB, Wadden TA, Fabricatore AN. Psychosocial and behavioral aspects of bariatric surgery. Obes Res 2005; 13: 639–648. 4 Encinosa WE, Bernard DM, Steiner CA, Chen CC. Use and costs of bariatric surgery and prescription weight-loss medications. Health Aff 2005; 24: 1039–1046. 5 National Institutes of Health. Bariatric Surgery for Obesity, Vol. 8. NIH Publication, 2008; pp 1–6. 6 Frezza EE. Six steps to fast-track insurance approval for bariatric surgery. Obes Surg 2006; 16: 659–663. 7 National Institutes of Health. Gastrointestinal surgery for severe obesity consensus statement. NIH Consensus Development Conference 1991; March 25–27; Washington DC, pp 1–20. 8 Harris RP, Helfand M, Woolf SH, Lohr KN, Mulrow CD, Teutsch SM et al. Methods Work Group, Third US Preventive Services Task Force. Current methods of the US Preventive Services Task Force: a review of the process. Am J Prev Med 2001; 20: 21–35. 9 Dixon-Woods M, Agarwal S, Jones D, Young B, Sutton A. Synthesizing qualitative and quantitative evidence: a review of possible methods. J Health Serv Res Policy 2005; 10: 45–53. 10 Kuwada TS, Richardson S, Chaar ME, Norton JH, Cleek J, Tomcho J et al. Insurancemandated medical programs before bariatric surgery: do good things come to those who wait? Surg Obesity Rel Dis 2011; 7: 526–530. 11 Jamal M, DeMaria EJ, Johnson JM, Carmody BJ, Wolfe LG, Kellum JM et al. Insurance-mandated preoperative dietary counseling does not improve outcome and increases dropout rates in patients considering gastric bypass surgery for morbid obesity. Surg Obes Relat Dis 2006; 2: 122–127. 12 Ochner CN, Puma LM, Raevuori A, Teixeira J, Geliebter A. Effectiveness of a prebariatric surgery insurance required weight loss regimen and relation to postoperative weight loss. Obesity 2010; 18: 287–292. 13 Liu RC, Sabnis AA, Forsyth C, Chand B. The effects of acute preoperative weight loss on laparoscopic Roux-en-Y gastric bypass. Obes Surg 2005; 15: 1396–1402. 14 Martin LF, Tan TL, Holmes PA, Becker DA, Horn J, Bixler EO. Can morbidly obese patients safely lose weight preoperatively? Am J Surg 1995; 169: 245–253. 15 Alami RS, Morton JM, Schuster R, Lie J, Sanchez BR, Peters A et al. Is there a benefit to preoperative weight loss in gastric bypass patients? A prospective randomized trial. Surg Obes Relat Dis 2007; 3: 141–145. 16 Alvarado R, Alami RS, Hsu G, Safadi BY, Sanchez BR, Morton JM et al. The impact of preoperative weight loss in patients undergoing laparoscopic Roux-en-Y gastric bypass. Obes Surg 2005; 15: 1282–1286. 17 Pekkarinen T, Mustajoki P. Use of very low-calorie diet in preoperative weight loss: efficacy and safety. Obes Res 1997; 5: 595–602. 18 Alger-Mayer S, Polimeni JM, Malone M. Preoperative weight loss as a predictor of long-term success following Roux-en-Y gastric bypass. Obes Surg 2008; 18: 772–775. 19 Colles SL, Dixon JB, Marks P, Strauss BJ, O’Brien PE. Preoperative weight loss with a very-low-energy diet: quantitation of changes in liver and abdominal fat by serial imaging. Am J Clin Nut 2006; 84: 304–311. 20 Fris RJ. Preoperative low energy diet diminishes liver size. Obes Surg 2004; 14: 1165–1170. 21 Van Nieuwenhove Y, Dambrauskas Z, Campillo-Soto A, van Dielen F, Wiezer R, Janssen I et al. Arch Surg 2011; 146: 1300–1305. 22 Foster GD, Makris AP, Bailer A. Behavioral treatment of obesity. Am J Clin Nutr 2005; 82: 230–235.

23 Samuel I, Mason EE, Renquist KE, Huang YH, Zimmerman MB, Jamal M. Bariatric surgery trends: an 18-year report from the International Bariatric Surgery Registry. Am J Surg 2006; 192: 657–662. 24 Solomon H, Liu GY, Alami RS, Morton J, Curet MJ. Benefits to patients choosing preoperative weight loss in gastric bypass surgery: new results of a randomized trial. J Am Coll Surg 2009; 208: 241–245. 25 Ali MR, Baucom-Pro S, Broderick-Villa GA, Campbell JB, Rasmussen JJ, Weston AN et al. Weight loss before gastric bypass: feasibility and effect on postoperative weight loss and weight loss maintenance. Surg Obes Relat Dis 2007; 3: 515–520. 26 Parikh M, Dasari M, McMacken M, Ren C, Fielding G, Ogedegbe G. Does a preoperative medically supervised weight loss program improve bariatric surgery outcomes? A pilot randomized study. Surg Endosc 2012; 26: 853–861. 27 Harnisch MC, Portenier DD, Pryor AD, Prince-Petersen R, Grant JP, DeMaria EJ. Preoperative weight gain does not predict failure of weight loss or co-morbidity resolution of laparoscopic Roux-en-Y gastric bypass for morbid obesity. Surg Obes Relat Dis 2008; 4: 445–450. 28 Mrad BA, Stoklossa CJ, Bich DW. Does preoperative weight loss predict success following surgery for morbid obesity? Am J Surg 2008; 195: 570–573. 29 Taylor EL, Chiasson PM, Perey BJ. Prediciting bariatric surgical outcomes: does preoperative weight gain correlate with less postoperative weight loss? Obes Surg 1995; 5: 375–377. 30 Eisenberg D, Duffy A, Bell R. Does preoperative weight change predict postoperative weight loss after laparoscopic Roux-en-Y gastric bypass in the short term? J Obes 2010; Available online at: http://www.hindawi.com/journals/jobes/ 2010/907097.html. Accessed June 24, 2011. 31 Becouarn G, Topart P, Ritz P. Weight loss prior to bariatric surgery is not a prerequisite to excess weight loss outcomes in obese patients. Obes Surg 2010; 20: 574–577. 32 Phan DC, Flanders KM, Sanford L, Shah PC, Brams DM. Preoperative weight loss improves outcome after bariatric surgery. Surg Obes Relat Dis 2005; 1: 282 (abstract 106). 33 Huerta S, Dredar S, Hayden E, Siddiqui AA, Anthony T, Asolati M et al. Preoperative weight loss decreases the operative time of gastric bypass at a Veterans Administration hospital. Obes Surg 2008; 18: 508–512. 34 Fujioka K, Yah E, Wang HJ, Li Z. Evaluating preoperative weight loss, binge eating disorder, and sexual abuse history on Roux-en-Y gastric bypass outcomes. Surg Obes Relat Dis 2008; 4: 137–143. 35 Carlin AM, O’Connor EA, Genaw JA, Kawar S. Preoperative weight loss is not a predictor of postoperative weight loss after Roux-en-Y gastric bypass. Surg Obes Relat Dis 2008; 4: 481–485. 36 Riess KP, Baker MT, Lambert PJ, Mathiason MA, Kothari SN. Effect of preoperative weight loss on laparoscopic gastric bypass outcomes. Surg Obes Relat Dis 2008; 4: 704–708. 37 Juńior WS, do Amaral JL, Nonino-Borges CB. Factors related to weight loss up to 4 years after bariatric surgery. Obes Surg 2011; 21: 1724–1730. 38 Aberle J, Freier A, Bushc P, Mommsen N, Beil FU, Dannheim V et al. Treatment with Sibutramine prior to Roux-en-Y gastric bypass leads to an improvement of metabolic parameters and to a reduction of liver size and operative time. Obes Surg 2009; 19: 1504–1507. 39 Benotti PN, Still CD, Wood GC, Akmal Y, King H, El Arousy H et al. Preoperative weight loss before bariatric surgery. Arch Surg 2009; 144: 1150–1155. 40 Blackburn G. Effect of degree of weight loss on health benefits. Obes Res 1995; 3: 211–216. 41 Van de Weijgert EJ, Ruseler CH, Elte JW. Long-term follow-up after gastric surgery for morbid obesity: preoperative weight loss improves the long-term control of morbid obesity after vertical banded gastroplasty. Obes Surg 1999; 9: 426–432. 42 Shai I, Schwarzfuchs D, Henkin Y, Shahar DR, Witkow S, Greenberg I et al. Weight loss with a low-carbohydrate, mediterranean, or low-fat diet. N Engl J Med 2008; 359: 229–241. 43 Tarnoff M, Kaplan LM, Shikora S. An evidence-based assessment of preoperative weight loss in bariatric surgery. Obes Surg 2008; 18: 1059–1061.

Supplementary Information accompanies the paper on International Journal of Obesity website (http://www.nature.com/ijo)

