Intravenous fluid and electrolytes are given to resuscitate the patient from losses sustained during surgery and to maintain homeostasis during periods when oral intake may not be possible. Clearly, the more major the surgery, the greater is the intravenous fluid requirement. However, the optimum fluid replacement strategy during and following surgery remains controversial. Many centers do not have a protocol for intraoperative fluid administration and continue to advocate the delivery of 3 liters of fluid and 154 mmol of sodium per day to postoperative patients [1]. However, such fluid prescription regimens are reported to be associated with a positive fluid balance in excess of 3 liters by the fourth postoperative day [2]. This excess input of sodium and water may be associated with increased postoperative complications and an adverse outcome [3, 4]. Nonetheless, the precise nature of the current fluid and sodium prescription practice is unclear, as is the interaction between the magnitude of surgery, perioperative fluid prescription, and the development of postoperative complications.

This study documents the current practice regarding the duration and volume of postoperative fluid and electrolyte administration in patients undergoing elective colonic or rectal resection. It also aims to compare the effects on the outcome of such intravenous fluid and sodium administration in patients undergoing modest colonic surgery versus major rectal surgery.

Materials and Methods

Patients

A series of 100 consecutive patients undergoing elective colorectal resection with primary anastomosis over a 10-month period (October 2000 to July 2001) on one colorectal service were included in the study. Patients undergoing complex pelvic reconstruction or resection without colonic anastomosis were excluded. Patients were classified as having undergone segmental colonic or rectal resection (Table 1). Total colectomy was classified as a rectal procedure to reflect the extent of surgery involved and the need for rectal dissection.

Table 1. Nature of the operation performed in 100 patients undergoing colorectal resection.
Full size table

Preoperative bowel preparation comprised two sachets of Picolax on the day prior to surgery with unlimited clear oral fluids up to midnight. Thereafter patients were kept fasting. A single dose of a parenteral prophylactic antibiotic was administered on induction of anesthesia. No set guidelines for postoperative intravenous fluid management were imposed, and all patients were treated in accordance with the established practice of the attending anesthetist and one of three colorectal surgeons.

Study Design

This study was approved by the Lothian Research Ethics Committee. Patients’ case notes were reviewed retrospectively. Demographic information was gathered together with preoperative anthropometric results, American Society of Anesthesiologists (ASA) grade, the Physiological and Operative Severity Score for the Enumeration of Mortality and Morbidity (POSSUM) score, and the operative indication [5].

Operative factors including the nature of the surgery, the duration of the surgery, and the intraoperative blood loss were identified. The quantity of intravenous fluid (including blood and blood products) and sodium replacement including antibiotic administration (which may contain a significant quantity of sodium) from midnight of the day of operation (day 0) to the fifth postoperative day (day 5) was recorded. The rate of postoperative intravenous fluid and sodium prescription was the mean of each individual’s daily intravenous fluid and sodium on the days they received intravenous fluids from day 1 to day 5.

Outcome measures included the duration of high dependency unit (HDU) care, time to cessation of intravenous fluids, postoperative hospital stay, in-hospital morbidity as defined in Appendix 1 [6], 30-day readmission rate, and 30-day mortality.

Statistics

Statistical analysis was performed using SPSS for Windows Release 11.0.0 (SPSS, Chicago, IL, USA). Continuous variables were compared with student’s t-test and the Mann-Whitney U-test. Categoric variables were compared using the ?2 and Fisher’s exact tests. Differences were considered significant at p < 0.05.

Results

The demographic profile, operative characteristics, and outcomes of patients are shown in Tables 2 and 3.

Table 2. Demographics, operative factors, and outcome of patients undergoing segmental colonic resection.
Full size table
Table 3. Demographics, operative factors, and outcome of patients undergoing rectal resection.
Full size table

The greatest input of both fluid and sodium occurred on the day of operation. Patients undergoing segmental colonic resection received a mean ± SEM intravenous fluid and sodium load of 3.6 ± 0.2 liters and 389 ± 22 mmol, respectively. Patients undergoing rectal resection received significantly more intravenous fluid and sodium on the day of surgery than did those having segmental colonic resection: a mean ± SEM of 4.6 ± 0.2 liters of fluid (p = 0.002) and 507 ± 34 mmol of sodium (p = 0.020).

By day 5 the mean cumulative total intravenous fluid and sodium inputs of patients having segmental colonic resections were 10.4 ± 0.5 liters and 874 ± 54 mmol, respectively. Once again, patients undergoing rectal resection received significantly more cumulative total intravenous fluid and sodium than patients having segmental colonic resection: 13.2 ± 0.6 liters of fluid (p < 0.001) and 1168 ± 67 mmol of sodium (p = 0.001).

Patients undergoing segmental colonic resections received a mean ± SEM daily rate of intravenous input of 1.8 ± 0.1 liters of fluid and 128 ± 8 mmol of sodium per day. This was significantly less than their counterparts undergoing rectal resection, who received 2.1 ± 0.1 liters of fluid (p = 0.031) and 155 ± 9 mmol of sodium (p = 0.031) per day.

The number and type of postoperative complications are shown in Table 4. The overall (segmental colonic and rectal cases combined) observed complication rate was 36%; the anastomotic leak rate was 5% and the 30-day mortality 1%.

Table 4. Postoperative complications.
Full size table

Outcome after Segmental Colonic Resection

Of the 44 patients undergoing segmental colonic resection, 17 (39%) developed a postoperative complication (Table 2). Those who developed postoperative complications had lengthier operations (p = 0.038) and were more likely to have required preoperative epidural analgesia (p = 0.008). There were no differences in the ASA grade or the POSSUM physiology or operative scores between patients who had had preoperative epidural analgesia and those who did not. There was one (2%) anastomotic leak and no 30-day mortality.

There was no statistically significant difference in mean intravenous fluid (p = 0.310) or sodium (p = 0.101) load on the day of surgery between patients who developed complications and those who did not. Patients who developed complications received a significantly greater mean ± SEM daily rate of intravenous sodium prescription: 149 ± 12 vs. 115 ± 10 mmol per day (p = 0.034) (Fig. 1a). However, this was not significant for the mean ± SEM daily rate of fluid input: 2.0 ± 0.1 vs. 1.7 ± 0.1 liters per day (p = 0.120) (Fig. 1b). The mean ±SEM cumulative intravenous fluid and sodium inputs over the first five perioperative days were 12.3 ± 0.8 liters and 1083 ± 91 mmol, respectively, for those with complications in contrast to 9.1 ± 0.5 liters of fluid (p = 0.001) and 742 ± 53 mmol of sodium (p = 0.001) in those without adverse outcomes.

Fig. 1.
figure 1

a. Rate of daily intravenous sodium administration after segmental colonic resection or rectal resection. Values are the means ± SEM. *p < 0.05 (unpaired student’s t-test). b. Rate of daily intravenous fluid administration after segmental colonic resection or rectal resection. Values are means ± SEM. Box with solid line = no complication; box with dotted line = complication.

Full size image

Outcome of Rectal Resection

Of the 56 patients undergoing rectal resection, 19 (34%) developed a postoperative complication (Table 3). There were four (7%) anastomotic leaks and one (2%) death at 30 days.

There was no statistically significant difference in mean intravenous fluid (p = 0.459) or sodium (p = 0.595) load on the day of surgery between patients who developed complications and those who did not. There was no statistically significant difference in the mean ± SEM daily rate of intravenous fluid [2.2 ± 0.1 vs. 2.0 ± 0.1 liters per day (p = 0.142)] and sodium [172 ± 20 vs. 146 ± 8 mmol per day (p = 0.166)] prescription for patients with complications and their counterparts without them (Fig. 1). There was no statistically significant difference in the mean ± SEM cumulative fluid and sodium inputs at day 5 [14.2 ± 1.1 vs. 12.7 ± 0.7 liters (p = 0.205) and 1275.0 ± 142.9 vs. 1113.0 ± 70.3 mmol (p = 0.257)] in patients who developed complications and those who did not.

Discussion

The work of Shires and colleagues in 1961 postulated a decrease in extracellular volume after surgery and recommended replacement of losses by additional fluid infusion. Since then, it has become routine practice for patients to be given large amounts of intravenous fluid after elective surgical procedures [7, 8]. The United Kingdom National Confidential Enquiry into Perioperative Deaths in 1999 recognized that errors in fluid and electrolyte management represented a significant cause of death [9]. However, there have been few contemporary studies documenting any practice in postoperative intravenous fluid replacement. The current data comprise an uncontrolled, retrospective review of a single center’s practice.

Fluid and electrolyte homeostasis in the postoperative patient is compromised by a reduced capacity to excrete excess sodium and water [2, 10, 11]. A common regimen for postoperative intravenous fluid prescription consists of 2 liters of 5% dextrose and 1 liter of 0.9% saline per day, which delivers 3 liters of water and 154 mmol of sodium. It has been suggested that such a policy may culminate in a significant positive fluid balance; and this excess administration of fluid may be associated with an adverse outcome and prolonged hospital stay [2, 3, 4].

The present data have shown that on the day of surgery the amount of fluid and sodium administration is greatly in excess of that of the regimen that comprises 2 liters of 5% dextrose and 1 liter of 0.9% saline daily. Patients undergoing both segmental colonic and rectal resection received approximately 35% of the 5-day cumulative total of intravenous fluid and 45% of the 5-day cumulative total of intravenous sodium on the day of surgery. Thereafter the patients received about 2 liters of fluid per day and 140 mmol of sodium: roughly 1 liter less of fluid but a similar load of sodium compared to the regimen of 2 liters of 5% dextrose and 1 liter of 0.9% saline. This proportionately more sodium than fluid in a number of patients is likely to stem from the use of sodium-rich colloid, crystalloid, and antibiotic solutions.

The benefits of restricting fluid to 2 liters per day and sodium to 77 mmol per day have been previously reported [2]. Brandstrup and colleagues’ randomized, controlled trial of perioperative fluid restriction after elective colorectal resection demonstrated fewer postoperative complications in the study arm; it was also noted that sodium input was reduced in that cohort [12]. Based on the current data it appears that it is the salt content as well as the fluid volume that requires reduction.Patients undergoing segmental colonic resection who went on to develop complications received a significantly greater cumulative total of intravenous fluid and sodium than those without complications. Such a clear difference was not identified in patients undergoing rectal resection. However, there was a stepwise trend for patients with or without complications across the two operative cohorts. Equally, it is possible that the magnitude of the surgery, in part, determines the effect of a relative fluid and sodium overload and outcome. Because colonic resection poses less of a physiological insult than rectal resection, the overall outcome in such patients may be more sensitive to the interplay between fluid and sodium input and the patient’s co-morbidity.

It could be argued that patients who developed complications required prolonged intravenous prescription. Therefore a larger cumulative sodium and fluid load would be an effect, rather than a cause, of postoperative complications. Indeed patients who developed complications after segmental colonic resection had more prolonged intravenous fluid administration, and a larger proportion of patients remained on intravenous fluid therapy up to day 5 than did patients without complications (59% vs. 15%; p = 0.006). However, analysis of the mean ± SEM daily input of the intravenous prescription on the first and second postoperative days after segmental colonic resection (when all patients remained on intravenous fluids) showed that those who developed complications had received a significantly higher rate of sodium administration on both the first (253 ± 37 vs. 170 ± 16 mmol/day; p = 0.024) and second (191 ± 21 vs. 143 ± 13 mmol/day; p = 0.042) postoperative days. Similarly, the daily mean ± SEM rate of intravenous fluid prescription was significantly higher on the second postoperative day for patients who developed complications than in those who did not (2.5 ± 0.1 vs. 2.1 ± 0.1 L/day; p = 0.027). These early differences in sodium and fluid prescription favor a causative relation, particularly in terms of sodium administration.

As noted previously, Lobo and colleagues reported the results of a prospective, nonblinded, randomized, controlled trial in patients undergoing segmental colonic resection comparing restriction of sodium to 77 mmol and fluid to 2 liters per day versus a conventional (2 liters 5% dextrose/1 liter 0.9% saline) fluid replacement strategy [2]. They noted significantly earlier discontinuation of intravenous fluids (day 4 vs. day 6) and postoperative discharge from hospital (day 6 vs. day 9) in their intervention arm. The control limb of the latter study had an approximate mean cumulative input of 16 liters of fluid and 1200 mmol of sodium by the fourth postoperative day, whereas the intervention cohort received 7.5 liters of fluid and 600 mmol of sodium.

Based on the present data, patients undergoing segmental colonic resection received a mean cumulative input of 10 liters of fluid and 849 mmol of sodium by the fourth postoperative day. Thus in the current series the doses of intravenous fluid and sodium administered were between those seen in the control and intervention arms of the study reported by Lobo et al. Similarly, outcomes in the present study (intravenous fluids discontinued on day 4, postoperative discharge on day 8) were between those of the control and the intervention arms of the study by Lobo and coworkers. Such a trend across studies tends to support the concept of a genuine relation between postoperative intravenous fluid/sodium administration and outcome.

The mechanisms by which excess perioperative intravenous fluid and sodium may mediate deleterious effects remain uncertain. An association between fluid overload, tissue edema, and reduced tissue oxygenation has been reported and may result in cardiorespiratory complications and compromised tissue healing [13]. It has also been reported that infusion of large amounts of 0.9% saline is associated with hyperchloremic metabolic acidosis. Such a derangement of acid-base status may impair cardiac contractility and reduce visceral perfusion [14].

Conclusions

An association between excess early salt and fluid prescription and adverse outcome after segmental colonic resection is highlighted by the present data. Although such a relation is supported by previous series, the optimum postoperative fluid regimen remains unclear. The present series supports a more conservative approach to perioperative sodium prescription in patients undergoing segmental colonic resection. However, further randomized controlled studies are required to determine if the large fluid and sodium load on the day of surgery can be reduced to beneficial effect.

Résumé

Un excès d’apports hydriques et sodés peut être responsable de complications postopératoires et d’évolution fâcheuse. Cependant l’importance d’un tel effet n’a pas été encore étudiée. Cette étude évalue les pratiques actuelles en ce qui concerne l’administration hydro-sodée après chirurgie colorectale et son rapport avec l’évolution postopératoire. Cent patients consécutifs ayant eu une résection élective colique (n = 44) ou rectale (n = 56) ont été inclus dans cette étude de cohorte rétrospective. Les données concernant le volume des apports hydro-sodés dans la perfusion par voie intraveineuse et l’administration des antibiotiques ont été enregistrées avec l’évolution clinique le jour de l’acte chirurgical et pour les quatre jours postopératoires consécutifs suivant l’intervention. La quantité moyenne (ET) des apports liquidiens et en sodium administrée le jour de l’intervention était plus grande après résection rectale qu’après résection colique [respectivement, 4.6 (0.2) vs. 3.6 (0.2) litres et 507 (34) vs. 389 (22) mmols (p < 0.05)]. La vitesse de perfusion moyenne (ETS) de liquide par jour d’eau et de sodium pour les quatre jours suivants a été plus grande après résection rectale qu’après résection colique [respectivement, 2.1 (0.1) vs. 1.8 (0.1) litres/jour et 155 (8.7) vs. 128 (8.0) mmols/jours (p < 0.05)]. Pour toutes les résections, il n’y avait aucune différence en ce qui concerne la quantité de liquide et de sodium administrée le jour de l’intervention chirurgicale avec ou sans complications postopératoires. Par contre, pour les quatre jours après, les patients ayant des complications après résection colique ont eu plus besoin de sodium en intraveineux que ceux qui n’en ont pas eu [149 (12) vs. 115 (10) mmols (p < 0.05)]. Ce besoin ne s’est pas manifesté de la même manière après résection rectale. La prescription postopératoire doit comporter approximativement deux litres d’apport hydrique et 140 mmols de sodium par jour. Les complications après résection colique, mais pas après résection rectale, sont associées à plus de besoins en sodium au début. Puisque la résection colique est moins agressive que la résection rectale au plan physiologique, l’évolution globale du premier groupe pourrait être plus sensible à la surcharge des apports hydro-sodés et à la co-morbidité du patient.

Resumen

Un exceso en la administraciın de agua y sodio puede verse asociado con complicaciones y una evoluciın clínica final adversa. Sin embargo, el efecto de la magnitud de la cirugía sobre tal asociaciın no ha sido debidamente analizado. El presente estudio fue diseñado para evaluar la práctica actual de la administraciın de líquido y de sodio por vía intravenosa luego de resecciın de colon y de recto y su relaciın con la evoluciın postoperatoria. Cien pacientes consecutivos sometidos a resecciın de colon (n = 44) o del recto (n = 56) fueron incluidos en un estudio retrospectivos de casos-cohorte. El promedio de líquido y de sodio administrado en el día de la operaciın apareciı mayor en los pacientes con cirugía rectal que en los de cirugía colınica [4.6 (0.2) vs. 3.6 (0.2) litros y 507 (34) vs. 389 (22) mmoles (p < 0.005)]. También lo fue la rata de administraciın durante los siguientes cuatro días [2.1 (0.1) vs.1.8 (0.1) litros/día y 155 (8.7) vs.128 (8.0) mmoles/día (p < 0.05)]. Considerando la totalidad de las resecciones, no se hallı diferencia en los volúmenes de sodio y de líquido administrados en el día de la cirugía entre los pacientes con o sin complicaciones postoperatorias. Pero en los siguientes cuatro días los pacientes con complicaciones luego de resecciın colınica registraron una mayor rata de administraciın de sodio que aquellos libres de complicaciones [149 (12) vs. 115 (10) mmols (p < 0.05)]. No se observı un patrın similar en los pacientes sometidos a resecciın rectal. La práctica corriente de prescripciın postoperatoria intravenosa de líquidos provee aproximadamente dos litros de líquido y 140 mmoles de sodio por día. Las complicaciones luego de reacciın rectal, pero no de resecciın colınica, aparecen asociadas con una mayor y más precoz administraciın diaria de sodio por vía intravenosa. Puesto que la resecciın colınica constituye un mayor trauma fisiolıgico que la resecciın rectal, el resultado global en el primer grupo puede ser más sensible a la interacciın entre la sobrecarga de líquido y de sodio y con la comorbilidad en el paciente que es sometido a cirugía.

Appendix: Definition of Complications [6]

Appendix 1.
Full size table