SEARCH

SEARCH BY CITATION

Keywords:

  • Gastric bypass;
  • laparoscopic;
  • meta-analysis;
  • morbid obesity

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

The aim of this meta-analysis was to assess the effects of laparoscopic vs. open gastric bypass for morbid obesity. A systematic review of the literature was undertaken to assess randomized controlled trials on laparoscopic and open gastric bypass for morbid obesity. Six randomized controlled trials involving a total of 422 patients were included. There were 214 patients in the laparoscopic group and 208 patients in the open group separately. Compared with open surgery, laparoscopic surgery for morbid obesity could significantly shorten hospital stays (WMD = −1.11 d, 95% confidence interval [CI][−1.65, −0.56]). However, laparoscopic surgery for morbid obesity showed higher re-operation (RR = 4.82, 95% CI [1.29, 17.98]) and longer surgical time (WMD = 28.00 min, 95% CI [7.84, 48.16]). There were no statistical differences in complication (RR = 0.84, 95% CI [0.64, 1.10]) and weight loss (WMD = 1.00 kg m−2, 95% CI [−0.79, 2.79]). The effects of laparoscopic and open gastric bypass for morbid obesity were basically the same except that laparoscopic had a shorter hospital stay and open surgery had a rate of fewer re-operations and shorter surgical time. Further high-quality, long follow-up period randomized controlled trials should be carried out to provide more reliable evidence.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

Morbid obesity is a serious health problem that occurs more and more frequently among younger age groups, which are usually associated with a series of comorbidities. According to the National Institutes of Health in 1991 (1), surgery is the only effective treatment of morbid obesity, indicated by a body mass index (BMI) of 40 or 35 with associated comorbidities.

The most commonly used surgical techniques are vertical banded gastroplasty and gastric bypass, the latter regarded by some as the ‘gold standard’ for the surgical treatment of obesity (2–4). Roux-en-Y gastric bypass (RYGBP) is a reliable surgical procedure with proven long-term efficacy (5) and safety (6) in the treatment of morbid obesity (3,4,7). The laparoscopic approach for treating morbid obesity has increased considerably in recent years because of the use of simple techniques involving low morbidity and mortality rates, such as gastric bypass (8,9). Reports in the literature have confirmed the technique's success with low complication rates and few recurrences (10,11). Since its advent, the laparoscopic technique has offered an alternative method and continues to gain acceptance among general (open) and laparoscopic (lap) surgeons. As the surgeon's experience and ability progress, other more complex techniques, such as gastric bypass or biliopancreatic diversions, are performed via laparoscopic surgery (12–14). The objective of this study is to evaluate the outcomes of laparoscopic vs. open gastric bypass for morbid obesity with weight change, surgical time, total duration of hospital stays, perioperative complication, mortality, re-operation and quality of life respectively.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) was used to conduct data extraction.

Study selection

We searched electronic databases from PubMed (1966–2009.5), the Cochrane Library (2009 Issue 2), EMBASE (1974–2009.5), Science citation index (1974–2009.5), The China Journal Fulltext Database (1994–2009.5), Chinese Scientific Journals Fulltext Database (1989–2009.5), Chinese Biomedical Literature Database (1978–2009.5), with the terms ‘obesity’ and ‘surg*’ and ‘operat*’, combination with the medical subject headings. Relevant articles referenced in these publications were downloaded from the databases. The related article function also was used to widen the search results. All abstracts, comparative studies, non-randomized trials and citations scanned were searched comprehensively. The flow chart of the literature search was shown in Table 1. We also hand-searched the reference lists of every primary study for additional publications. Further searches were done by reviewing abstract booklets and review articles. Trials were included irrespective of the language in which they were reported.

Table 1.  Baseline characteristics of included studies
TrialsYearSample size (n)CountryFollow-up (months)Age (years)Sex (male : female)BMI (kg m−2)
LapOpenLapOpenLapOpen
  1. BMI, body mass index; lap, laparoscopic; NR, not reported.

Sundbomand M and Gustavsson S 2004 (16)200425/25Sweden1237 (19–54)38 (24–54)2 : 233 : 2244 (36–54)45 (34–54)
Westling A and Gustavsson S 2001 (17)200122/21Sweden>1236 ± 936 ± 9NRNR42 ± 442 ± 4
Luján JA et al. 2004 (18)200453/51Spain2337 (18–64)38 (20–63)10 : 4313 : 3848.53 (36–78)52.20 (37–80)
Nguyen NT et al. 2001 (19)200136/34USA<1(41 ± 8)(43 ± 8)34 : 228 : 648 ± 550 ± 5
Puzziferri N et al. 2006(20)200659/57USA39 ± 847 ± 750 ± 856 : 351 : 648 ± 549 ± 6
Fisher BL 2004 (21)200419/20USANRNRNRNRNRNRNR

Data extraction

Each study was critically reviewed by two researchers for reliability of our meta-analysis. Only randomized controlled trials on laparoscopic vs. open gastric bypass for morbid obesity were included in the meta-analysis. Two researchers extracted data separately and if there was controversy, it was confirmed by the third researcher.

Inclusion criterion

Patients with a BMI of 40 to 60 kg m−2 undergoing evaluation for bariatric surgery, aged 21 to 60 years, and who had failed at all previous medical interventions for weight loss were included.

Exclusion criteria

Patients who had previous bariatric surgery, previous gastric surgery, a large abdominal ventral hernia, a history of deep venous thrombosis or pulmonary embolism, and severe cardiovascular, respiratory, hepatic, or renal disease were excluded.

Statistical analysis

We summarized available data from all trials reporting results. Results for continuous outcomes as weighted mean difference or standard mean difference and dichotomous outcomes as risk ratios (RR) with 95% confidence intervals (CI) were expressed. All statistical analyses were performed with Review Manager (RevMan version 5.0). We used the chi-squared statistic to assess heterogeneity between trials and the I2 statistic to assess the extent of inconsistency. If there was a significant heterogeneity, a random-effects statistical model was used to confirm the case results. A fixed-effect model for calculations of summary estimates and their 95% CIs was also applied, unless there was a significant heterogeneity. Subgroup analysis was intended to explore important clinical differences among trials.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

From Fig. 1 we can see the flow chart studies from the initial results of publication searches to the final inclusion. Six trials of laparoscopic vs. open gastric bypass for morbid obesity totalling 422 patients were retrieved from the electronic databases. There were 214 patients in the laparoscopic group and 208 in the open group. Standard deviations were not reported in the majority of studies, where necessary they were estimated either by means of ranges or P-values. Characteristics of each trial were given in Table 1. The outcomes extracted from these trials were shown in Table 2. The methodological quality of the included studies was assessed using the Cochrane handbook (15) in Table 3.

image

Figure 1. Flow diagram of trial selection.

Download figure to PowerPoint

Table 2.  Methodological quality of included studies
StudiesRandomizationAllocation concealmentBlindingIncomplete outcome dataITT Analysis
  1. NR, Not reported; ITT, intention to treat.

Sundbomand M and Gustavsson S 2004 (16)Not statedSealed envelopeDouble blindYesNot stated
Westling A and Gustavsson S 2001 (17)Not statedSealed envelopeNot statedYesNot stated
Luján JA et al. 2004 (18)Computer generatedSealed envelopeNot statedYesNot stated
Nguyen NT et al. 2001 (19)Not statedSealed envelopeNot statedYesNot stated
Puzziferri N et al. 2006 (20)Not statedSealed envelopeNot statedYesNot stated
Fisher BL 2004 (21)Not statedNot statedNot statedYesNot stated
Table 3.  Outcome variables
TrialsBMI change (kg m−2)Surgical time (min)Re-operationStay (d)
LapOpenLapOpenLapOpenLapOpen
  1. BMI, body mass index; lap, laparoscopic; NR, not reported.

Sundbomand M and Gustavsson S 2004 (16)1515150 (110–265)85 (60–150)106 (4–14)6 (3–7)
Westling A and Gustavsson S 2001 (17)14 ± 313 ± 3235 (165–390)100 (70–150)914 ± 0.86 ± 3.8
Luján JA et al. 2004 (18)NRNR186.4 (125–290)201.7 (129–310)NRNR5.2 (1–13)7.9 (2–28)
Nguyen NT et al. 2001 (19)NRNR230 ± 46202 ± 40NRNR3 (median)5 (median)
Puzziferri N et al. 2006 (20)NRNRNRNR21NRNR
Fisher BL 2004 (21)NRNRNRNRNRNR2.5 (2–4) (SD = 0.70)3.5 (3–7) (SD = 1.10)

Weight loss

Four studies reported weight loss. There was no statistically differences in weight loss (weighted mean difference (WMD) = 1.00 kg m−2, 95% CI [−0.79, 2.79]) (Fig. 2a). The weight loss expressed in decrease in the percentage of excess weight loss (%EWL) and BMI.

image

Figure 2. (a) BMI change, (b) surgical time and (c) hospital stay. BMI, body mass index; CI, confidence interval; lap, laparoscopic.

Download figure to PowerPoint

Surgical time

Still no heterogeneity existed among trials (I2 = 0%, P = 0.94). Open surgery for morbid obesity was associated with shorter surgical time (WMD = 28.00 min, 95% CI [7.84, 48.16]) (Fig. 2b).

Hospital stay

Two studies (18,22) reported post-operative length of hospital stay. Laparoscopic surgery for morbid obesity was associated with significantly shorter hospital stays (WMD = −1.11 d, 95% CI [−1.65, −0.56]) (Fig. 2c).

Re-operation

No heterogeneity existed among trials (I2 = 0%, P = 0.47), there was significant difference in re-operation between laparoscopic and open surgery (RR = 4.82, 95% CI [1.29, 17.98]) (Fig. 3a).

image

Figure 3. (a) Re-operation and (b) complications. CI, confidence interval; lap, laparoscopic.

Download figure to PowerPoint

Complications

Statistical heterogeneity was basically good among studies (I2 = 0%, P = 0.77). Six studies (17–21) reported complications. We used fixed effects models. There were no differences between the two groups. (RR = 0.84, 95% CI [0.64, 1.10]) (Fig. 3b). Individual complications reported in each study included are shown in Table 4.

Table 4.  Reported complications of included studies
TrialsComplications
Laparoscopic (n)Open (n)
  1. GJ, gap junctions; UGIH, upper gastrointestinal hemorrhage; LGIH, lower gastrointestinal hemorrhage.

Sundbomand M and Gustavsson S 2004 (16)Intra-abdominal bleeding (n = 3)Intense in traluminal bleeding (n = 2)
Westling A and Gustavsson S 2001 (17)Pulmonary embolism (n = 1) Stomal ulcer (n = 1) Stenosis of GJ (n = 1) Infection (n = 0) Incisional hernia (n = 0)Pulmonary embolism (n = 0) Stomal ulcer (n = 0) Stenosis of GJ (n = 0) Infection (n = 3) Incisional hernia (n = 1)
Luján JA et al. 2004 (18)Intestinal subocclusions (n = 3) Asymptomaticleaks (n = 2) Intra-abdominal bleeding (n = 2) UGIH (n = 2) LGIH (n = 1) Thrombophlebitis (n = 1) Stenosis of the gastroenteroanastomosis (n = 1)Subphrenic abscesses (n = 4) UGIH (n = 3) Wound infections (n = 4) Respiratory infections (n = 3) Evisceration (death) (n = 1)
Nguyen NT et al. 2001 (19)Respiratory failure (n = 0)Respiratory failure (n = 1) Pulmonary embolism (n = 1)
Puzziferri N et al. 2006 (20)Incisional hernia (n = 3) Anaemia (n = 8) B-12 deficiency (n = 3) Chronic nausea/vomiting (n = 3) Chronic abdominal pain (n = 2) Marginal ulcer (n = 0) Small bowel obstruction (n = 2) Cholecystectomy (n = 12)Incisional hernia (n = 22) Anaemia (n = 3) B-12 deficiency (n = 6) Chronic nausea/vomiting (n = 2) Chronic abdominal pain (n = 1) Marginal ulcer (n = 1) Small bowel obstruction (n = 1) Cholecystectomy (n = 2)
Fisher BL 2004 (21)Not reportedNot reported

Mortality

Three trials (17–19) reported that patients died in the post-operative period. In Luján JA et al. 2004, two in the LGBP group, one case not related to the surgical procedure and one in the OGBP group died of broncho-aspiration in the re-operation for evisceration secondary to coughing on the second post-operative day. In Sundbomand M and Gustavsson S 2004, one patient died 11 months after operation from metastatic breast cancer; she received oncological treatment and conventional terminal care, which were not affected by her previous RYGBP. In Westling A and Gustavsson S 2001, one patient in the laparoscopic group died of multiple organ failure 1 week post-operatively.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

Gastric bypass is the operation recommended by the National Institute of Health for treatment of morbid obesity because of its low morbidity and mortality rates and excellent long-term results regarding weight loss. Witgrove was the first to perform it via the laparoscopic approach in 1994 (14). Since then, others have performed this procedure with different techniques such as creation of a gastroenteroanastomosis, Roux-en-Y positioning of the loop in relation to the colon and stomach, the length of this loop and use of anentero-enterostomy, the great majority reporting good results (13,22–24).

Results for meta-analysis showed that laparoscopic surgery had a shorter hospital stay, higher re-operation rates, while the open group needed less surgical time. There were no differences in complications and weight change.

Length of hospital stay results from the studies included in this meta-analysis showed that laparoscopic significantly reduced length of hospital stay as compared with open by 0.48 d. The results are due to the small surgical wounds and also may be due to the fact that mobilization following laparoscopic is improved, thereby facilitating recovery and subsequent discharge from hospital. Why this occurs may be related to wound pain, infection and ileus formation, all of which impair a patient's mobility following an operation. Although this is a small reduction for a single case, reducing length of stay by this amount for every patient undergoing gastric bypass is likely to make a significant difference to cost of post-operative care. Needless to say, this is a link that must be further researched in the future.

Results of surgical time showed that, although laparoscopic took longer than Open, this was not statistically significant. The concept of a laparoscopic procedure taking longer than its open equivalent is not surprising, because the laparoscopic procedure was more complex than the open procedure.

Four trials (16–18,21) reported weight loss for two groups. Data for weight loss was hard to pool. Therefore, we can only take the descriptive analysis. All of the trials said it was similar in both groups. Whether the procedure offers a long-lasting effect is an important question for concern. More importantly, weight gain should also be a focus.

The mortality rate, in both groups, was very low, with no differences between the two groups. However, the tragic mortality in our study could not only be linked to this type of operation, but could be found in any type of surgeries.

When questioned in the trials (17,21) post-operatively, the vast majority of our patients were very satisfied with their weight loss and their quality of life had improved dramatically. We did not find a difference between patients who had had an open or a laparoscopic procedure.

Most of the complications were the same in the two groups in Table 4. However, wound infection and ventral hernia tended to be more probable following open surgery, and stricture of the tunnel through the mesocolon occurred more following laparoscopy.

One study (16), of the nine included studies, offered adequate descriptions of the randomization process. Five studies (16–20) offered allocation concealment, and only one study (16) reported blinding. Meanwhile, the randomization process in the study (16,17,19–21), allocation concealment in the trial (21) and blinding in the study (17–21) were not stated, which would yield selection bias and performance bias. Furthermore, none of the studies reported intention-to-treat analysis, which would yield attrition bias.

Our meta-analysis also had its limitations. The heterogeneity of some variables in this study is worthy of comment. First, some of the included studies (19,21) focused only on single aspects. Second, the different operative methods were performed by different surgeons in different countries. Different learning curves may influence the difference between the two procedures. Third, only some of the included studies (16–18,20) had 12 months or longer follow-up. However, expert opinion suggests that follow-up outcomes should be considered for more than 5 years. Fourth, some data should be assessed using weighted mean differences, or we can only take the descriptive analysis. Finally, we were unable to assess effects of baseline patient characteristics on estimates because we could not access the individual patient data.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

There was no significant difference between laparoscopic and open gastric bypass in weight loss. The limited evidence suggests that laparoscopic had shorter hospital stays and open surgery had a smaller re-operation rate and shorter surgical time. Further high-quality, long follow-up period randomized controlled trials should be carried out to provide more reliable evidence.

Conflict of interest statement

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

There is no financial support or relationships that may pose conflict of interest in this study paper.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References

Thank you very much to the co-authors and all authors, who have contributed significantly. All authors are in agreement with the content of the manuscript. The copyright will be sent to your Journal.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflict of interest statement
  9. Acknowledgements
  10. References