Perioperative intravenous fluid prescribing: a multi-centre audit

Harris B, Schopflin C, Khaghani C, Edwards M.

perioper med (lond). 2015; 4:15.

This multicentre audit reviewed the records of 431 patients who had undergone major elective surgery. The authors sought to determine how well IV fluids were prescribed both intraoperatively and perioperatively when compared with current guidance. On the positive side, they found that patients were allowed oral fluids early; however, there was huge variation in the volume of fluid given intraoperatively, an excess of sodium and fluid overall often being given postoperatively, and potassium supplementation was inadequate.

The fundamental goal of fluid management is to maintain the fluid distribution and electrolyte balance of a patient where their own systems are unable to do so. In the surgical setting, both inadequate and excessive fluid therapy have been linked with poorer outcomes; this has led to increased emphasis on optimal fluid management, and guidelines have emerged highlighting areas of risk. Despite this recognition, the evidence base surrounding perioperative fluid management remains inconsistent. Compounding this are entrenched attitudes to practice, as well as recognition that deviations from accepted practice are appropriate at times. Several studies have suggested that despite the guidelines, prescribing remains suboptimal. The basis of this study was to determine to what extent practice adhered to the guidance.

This audit was carried out across five hospitals in southern England, and the intention was to analyse the notes of 100 patients per hospital, and 150 in the largest centre. In practice, not all of the notes were available, resulting in 431 of 550 cases being reviewed. The population was adults undergoing elective surgery. Minor procedures, procedures performed under regional anaesthetic, and day-case surgeries were excluded. To minimise bias, the first 100 eligible patients for each year were included. Demographic information including weight, age and operation type were collected, as was information on the type of fluid administered and the quantities of electrolytes. Fluid balance was calculated from input/output charts, and serum electrolyte measurement was reviewed where available. Data were reviewed until day 3 post-op, the day after IV fluids were ceased, or the day of discharge – whichever occurred first.

Of the cases reviewed, 43% were orthopaedic, 23.5% were upper gastrointestinal (GI), the remaining 33.5% were lower GI, gynaecological or urological. The most common fluids prescribed were balanced crystalloids, particularly in the intraoperative period, while dextrose saline and 0.9% saline began to increase in use postoperatively. Nearly all patients received some fluid intraoperatively, with the most common procedures being laparoscopic right hemicolectomy, total hip replacement and total knee replacement. There was marked variation in the intraoperative volume of fluid given, that was not linked to bodyweight. Postoperatively, around 50% received IV fluids on day 1, dropping to around 25% on day 2 – there was no correlation between bodyweight and volume administered. For those patients on IV fluids on day 1, around half the patients received excess sodium and water, while almost all received less than maintenance potassium. The incidence of hyponatraemia increased progressively after surgery, and was most marked in those given dextrose saline. The incidence of hypokalaemia also increased progressively; 4% were hypokalaemic on day 1, 6% on day 2 and 10% on day 3, up from 1.5% preoperatively.

The authors point to the drop off in patients on IV fluids as evidence that enhanced recovery principles were being followed – allowing early oral fluids. In those patients continuing on fluids, electrolyte disturbances became increasingly common, and replacement was not being optimised. They also note the variation in volumes received intraoperatively as a reflection of erratic application of guidance.

Strengths of the study were:

• its size 
• the common procedures included 
• the wide geographical area covered 
• the fact that its findings were consistent with previous studies
• the fact that it examines the entire perioperative period.

Limitations were:

• the majority of notes were paper-based and often of poor quality 
• information about complications was based on discharge summaries and the anaesthetic chart which may not have been accurate
• the length of operation was not recorded – precluding a measure of volume of fluid per kilogram per hour
• the perioperative risk scores were not matched.

Suggestions to improve the way fluids are prescribed included education – particularly for more junior doctors, as ward-based prescribing and awareness of electrolyte imbalance was not done effectively, and hypokalaemia is linked to a prolonged recovery. The observed hyponatraemia was most likely due to the physiological injury due to surgery, but there may be an effect from dextrose saline.

This study provides an important snapshot of how fluid is being prescribed, and shows it to be sub-optimal in practice. The authors emphasise the need for regular audit cycles to examine how well any centre adheres to best practice. This study is particularly important as it provides a much greater context for how patients are managed, including data examining the ward-based setting – data collection has typically been based off anaesthetic charts, which give a more easily quantified, yet more limited view of practice.

Access Article Abstract