COST-EFFECTIVENESS OF THE USE OF ACERTO PROTOCOL IN MAJOR DIGESTIVE SURGERY

ABSTRACT - BACKGROUND: Hospital costs in surgery constitute a burden for the health system in all over the world. Multimodal protocols such as the ACERTO project enhance postoperative recovery. OBJECTIVE: The aim of this study was to analyze the hospital costs in patients undergoing major digestive surgical procedures with or without the perioperative care strategies proposed by the ACERTO project. METHODS: Retrospective data from elective patients undergoing major digestive surgical procedures in a university hospital between January 2002 and December 2011 were collected. The investigation involved two phases: between January 2002 and December 2005, covering cases admitted before the implementation of the ACERTO protocol (pre-ACERTO period), and cases operated between January 2006 and December 2011, after implementation (ACERTO period). The primary outcome was the comparison of hospital costs between the two periods. As secondary end point, we compared length of stay (LOS), postoperative complications, surgical-site infection (SSI) rate, and mortality. RESULTS: We analyzed 381 patients (239 of the pre-ACERTO period and 142 of the ACERTO period) who underwent major procedures on the gastrointestinal tract. Patients operated after within the ACERTO protocol postoperative LOS had a median of 3 days shorter (p=0.001) when compared with pre-ACERTO period [median (IQR): 10 (12) days vs. 13 (12) days]. Mortality was similar between the two periods. Postoperative complications risk, however, was 29% greater (RR: 1.29; 95%CI 1.11-1.50) in the pre-ACERTO period (p=0.002). SSI risk was also greater in pre-ACERTO period (RR: 1.33; 95%CI 1.14-1.50). Costs (mean and SE) per patients were R$24,562.84 (1,349.33) before the implementation and R$19,912.81 (1,459.89) after the ACERTO protocol (p=0.02). CONCLUSION: The implementation of the ACERTO project in this University Hospital reduced the hospital costs in major digestive procedures. Moreover, the implementation of this modern perioperative care strategy also reduced postoperative complications, SSI risks, and LOS.

This study analyzes costs using Brazilian reais and we hope these figures could influence surgeons to adopt the ACERTO protocol of perioperative care in major digestive procedures. These would help not only patients but also the costs of health system.

Central Message
This study focused on the hospital costs reduction using the ACERTO project of perioperative care in major digestive procedures. Costs per patients were reduced by approximately 20% as well as the postoperative complications and surgicalsite risks.
The figure shows the difference of the mean cost per patient in Brazilian reais before and after the implementation of the ACERTO project.

INTRODUCTION
Multimodal protocols became known around the world after the introduction of the so-called "fast-track surgery" Kehlet and Wilmore in 1980s 13 . The authors defined fast-track surgery as a multimodal strategy to care using a combination of epidural or regional anesthesia, minimally invasive techniques, optimal pain control, aggressive postoperative rehabilitation, and postoperative early enteral or oral nutrition 21 .
The central idea was to reduce the stress response and to abbreviate the recovery after surgery. For this, the ERAS group included more protocols such as the shortening of preoperative fast to the fast-track strategy and published various guidelines to approach with different surgical procedures 11,14 .
In Brazil, the ACERTO (Aceleração da Recuperação Total Pós-Operatória -Postoperative Enhanced Total Recovery) multimodal protocol was launched in 2005 and was reported first in 2006 1 . A guideline for the implementation of the ACERTO project was published in 2017 7 after various articles assured the reduction of important end points such as postoperative length of stay (LOS), postoperative complications, and mortality when compared with traditional care 1,4,6 .
Hospital costs in surgery represent a burden for the health system all over the world 2,11,19,22 . In this context, the use of ERAS strategies of perioperative care has consistently shown that multimodal protocols can reduce costs and improve costeffectiveness 12,18,20 . The reduction in hospital costs can be seen not only in major operations but also in hernioplasties in our hospital through the modification from traditional care to the ACERTO protocol 16 . However, until now, we did not have an analysis of costs in major procedures using the ACERTO protocol.
Furthermore, we could not find this analysis in our currency in other studies. We then hypothesized that, as we have initially found in hernioplasties, the cost-benefit should be greatly reduced in major operation with the new implemented multimodal protocol of perioperative care in our university hospital.
Thus, the aim of this study was to analyze the hospital costs in patients undergoing major digestive surgical procedure with or without the perioperative care strategies proposed by the ACERTO project. The investigation involved two phases: between January 2002 and December 2005, covering cases admitted before the implementation of the ACERTO protocol (pre-ACERTO period), and the other, with cases operated between January 2006 and December 2011, after its implantation (ACERTO period). Table 1 shows the protocols established by ACERTO protocol and the conventional procedures that had been applied before its implementation in the infirmary of the hospital.

METHODS
All patients underwent nutritional assessment by the subjective global assessment (SGA), as previously described 1 . In summary, patients bearing score A were considered eutrophic and if scoring B or C they were considered malnourished.
The main end point was the daily total cost of hospitalization, comparing the two periods studied according to the method described below. The hypothesis, formulated prior to data collection, was that patients undergoing the ACERTO perioperative care would have lower daily total costs due to reduced postoperative complications, surgical-site infection (SSI) rate, and a shorter LOS. Accordingly, as a secondary end point, we compared LOS, postoperative complications, SSI rate, and mortality in both periods. Postoperative complications and SSI were defined according to the criteria proposed by Mangram et al. 15 .

Cost analysis
The primary outcome of the study was the difference in hospital costs between the two periods. We used the costs accrual method according to NBCT 16.11 -Public Sector Cost Information System 5, 16 . This method allows an indirect calculation of daily cost of the patients as follows. To obtain the average cost of hospitalization per patient per day, we divided the total costs of hospitalization in the infirmary of Surgery Clinics by the patient/day annual average. As for the calculation of the average cost per number of hospitalizations, we divided the total costs of admissions to the Surgical Clinics by the number of hospitalizations performed in each period. Finally, the value of the average cost of hospitalization per night consisted of dividing the total costs of admission to the Surgical Clinic by the number of daily rates in the period. For the purposes of calculating the average cost of hospitalization at the surgical clinics of the HUJM, we used the following data: (1) product output report by sector issued by the MV 2000 inventory control system; (2) laboratory and image examination report issued by the MV 2000 exam billing system; (3) authorizations for hospital admission (AIH) movement report -reduced files and rejected AIH issued by the DataSUS/Tabwin system; (4) personnel data sheet for public employees provided by the HUJM human resources unit; (5) personnel data from the servants of the single legal regime (RJU) provided by the HUJM expense settlement and payment unit; (6) work schedule available on the HUJM website; (7) information on the number of equipment in the operating room provided by the head of that unit; (8) data of the clinical engineering contract, as well as footage of the HUJM hospital areas made available by the logistics and infrastructure division; and (9) information on accommodation costs obtained by the hospital accommodation indicators monitoring panel and made available by the hospital accommodation unit. We thus obtained the value of R$1,442.86 for the daily cost of a patient operated on our infirmary.

Statistical analysis
We planned to do an intention-to-treat analysis, meaning the comparison of the two periods disregarding if, especially in the second period, the patient have received or not received the ACERTO protocol. The normality of the continuous variables was assessed with the Kolmogorov-Smirnov test, and the homogeneity of their variances with the Levene test. To compare daily cost and the length of hospital stay, we used the Student's t-test accordingly. All other continuous variables were compared using the Mann-Whitney U test. We expressed all continuous data as median and interquartile range (IQR) or as mean and standard error (SE) accordingly. We analyzed categorical variables (i.e., surgical complications, SSI, and deaths) using the chi-square test. We adopted a value of p<0.05 as the statistical significance threshold. As a measure of the association strength, we calculated the relative risk (RR), with 95% confidence interval (95%CI). All calculations were performed using the SPSS statistical package version 20.0.

RESULTS
During the period of the study, 4,071 elective procedures were carried out in the hospital (1,805 patients, 44.3% in the pre-ACERTO period; and 2,266 patients, 55.7% in the ACERTO period). Of these, 2,014 patients were submitted gastrointestinal procedures. We excluded 1,633 patients due to minor procedures such as cholecystectomies, hernioplasties, and anal procedures (n=1452); emergency or urgency operations (n=106); and missing data (n=75). We then analyzed 381 patients who underwent major digestive procedures. Table 2 shows the demographic and clinical variables of these patients according to the two periods of the study. For this study, esophagectomy, any surgical procedure to megaesophagus, total or partial gastrectomy, gastroenteric anastomosis, biliodigestive anastomosis, gastro-or duodeno-pancreatectomy, partial pancreatectomy, colorectal resection, and colostomy closures were considered the major procedures.

Length of Stay
Patients operated after the implementation of the ACERTO protocol had a median (IQR) postoperative LOS of 3 days shorter (p=0.001) when compared with those from the pre-ACERTO period [10 (12) days vs. 13 (12) days].

Hospital costs
The mean (SE) cost of each patient after the implementation of the ACERTO protocol was reduced by almost 20% when compared with before the implementation. The mean reduction in our hospital with 100% patients financed by SUS (Sistema Único de Saúde -Brazilian Public Health System) and submitted to major digestive operations was R$4,650.03. Costs per patients were R$24,562.84 (1,349.33) before the implementation and R$19,912.81 (1,459.89) after the use of the ACERTO protocol (p=0.02) (Figure 1).

DISCUSSION
Our findings showed that the implementation of a multimodal protocol of perioperative care such as the ACERTO protocol reduces costs in major digestive surgical procedures. These findings probably are associated with the concomitant decrease of postoperative complications and LOS. This is the first study analyzing the reduction of costs with the ACERTO project in major operations using Brazilian currency. Various studies which compared costs with or without ERAS multimodal protocol showed the same with international currencies such as Euros, British Pounds, or Swiss Francs. 12,18,20 The overall results seem to agree that changing traditional to modern perioperative care may decrease costs and postoperative morbidity.
We used a strategy of cost analysis using an indirect method that assessed daily total hospital charges as previously described 16,19 . The administrative core of the hospital was involved and produces data from the various units of costs, which ail the base of calculations. Our first study with this method was published years ago showing that the type of perioperative care may modify costs in hernioplasties. In the present study, we endorse our first findings showing this time that the reduction in Brazilian Reais is also finding in major operations.
The reduction of LOS and postoperative complications, especially SSI, had already been reported with the ACERTO perioperative care in elective procedures 1,3,8 . By changing the perioperative nutritional approach by means of a protocol rather than the staff criteria probably had an important role in these results. Moreover, the decrease of preoperative fasting to 2 h, only used after the implementation of the new protocol, may have also contributed to the better results 9 . Decreasing the fast time before and after the surgical procedure reduced the organic response to trauma and decrease not only the LOS but also postoperative complications in various studies 4,9 . Nutritional attention is mandatory and should be implemented in these patients according to many guidelines based on evidence 7,10,14 .
Pimento et al. have recently reported that healthcare costs can be reduced by the implementation of nutrition intervention for patients with gastrointestinal cancers 17 . We agree with that, since the implementation of the ACERTO project changed the nutritional approach in our patients from a staff-oriented perspective to a protocol-oriented protocol, which was absorbed by the multidisciplinary staff of the hospital including not only surgeons but also nurses, dietitians, and physiotherapists.
However, our findings have limitations. First, hospital daily charges were assumed to be an accurate surrogate for hospital costs. More accurate methods may be available, but we were unable to access. Second, as a retrospective study, there is a time-dependent bias, which can lead to overestimating or underestimate the financial impact 19 . Although all patients were from the same public health system and were operated on a single center, and with the same surgeons, the data may also have bias due to the different number of patients undergoing major operations between the two periods.
The new public hospitals in various cities of Mato Grosso State along with new surgery residency programs may have contribute to a decrease in the number of major operations during the second period of the study. We believe that patients bearing surgical oncological digestive diseases once referred to our University hospital were operated on in new hospitals during the last period. However, we compared only major procedures in the two periods. Due to this, we considered our findings as appropriate to be compared though the total number of procedures in the two periods were uneven. A decrease in costs using ERAS protocol by comparing two periods as we used in this study has been reported as well, which is consistent with our findings 12,20 .

CONCLUSION
The implementation of the ACERTO project in the University hospital, including only patients bearing the Brazilian public health system insurance, reduced the hospital costs in major digestive procedures. Moreover, the implementation of this modern perioperative care strategy also reduced postoperative complications, SSI, and LOS.