Fluid balance from intake and output

Clinical background

Fluid balance from intake and output is neither a risk score nor a diagnostic model, but a simple accounting method for following the cumulative net fluid balance in patients in whom volume status and fluid distribution are difficult to judge clinically. This applies primarily to critically ill patients in intensive care and to perioperative patients with large fluid shifts. The need for a structured intake and output chart arises because clinical signs of volume status (oedema, jugular venous filling, capillary refill, lung auscultation) correlate poorly with actual volume status in the critically ill patient, and because a cumulatively positive fluid balance has been shown to be an independent risk marker for increased mortality.

In a prospective multicentre study of 618 adult patients with acute kidney injury in intensive care, fluid overload was defined as an increase in body weight of more than 10 per cent from baseline. Patients meeting this criterion had significantly higher 60-day mortality, and at the start of dialysis the adjusted odds ratio for death associated with fluid overload was 2.07 [1]. A cumulatively positive balance over the first three days in intensive care was moreover an independent risk factor for 28-day mortality in a separate prospective multicentre study of 2,526 patients, of whom 1,172 developed acute kidney injury [2].

Calculating the fluid balance from intake and output

The net balance is calculated as the difference between all measured fluid inputs and all measured and estimated fluid losses over the period in question:

$$\text{Net balance} = (V_{\text{iv}} + V_{\text{blood/colloid}} + V_{\text{oral/ent}}) - (V_{\text{urine}} + V_{\text{drain}} + V_{\text{stool}} + V_{\text{insensible}})$$

where $V_{\text{iv}}$ is intravenous fluid and drugs, $V_{\text{blood/colloid}}$ is blood and colloid products, $V_{\text{oral/ent}}$ is oral or enteral intake, $V_{\text{urine}}$ is urine output, $V_{\text{drain}}$ is drainage, nasogastric aspirate and vomit, $V_{\text{stool}}$ is stool and other losses, and $V_{\text{insensible}}$ is the estimated insensible losses.

The calculation is usually performed per 24 hours and then summed into a cumulative balance over the stay. Insensible losses, comprising perspiration from the skin and losses from the airways, are estimated at rest at approximately 500 to 800 mL/day in an afebrile patient at room temperature. With fever the losses increase by approximately 10 per cent per degree above 37 °C, and with tachypnoea airway losses increase further. These losses are not measured but estimated, and this is the only variable in the formula that does not rest on an actual measurement.

The evidence base for the clinical use of cumulative fluid balance as a basis for decision comes from prospective observational studies in intensive care cohorts. Bouchard et al. [1] used the increase in body weight as a measure of fluid accumulation in a multicentre cohort of 618 patients with acute kidney injury and found that accumulation exceeding 10 per cent of baseline weight was independently associated with increased mortality and poorer recovery of renal function. Wang et al. [2] showed in a multicentre study of 2,526 intensive care patients that the cumulative fluid balance over three days was an independent risk factor for 28-day mortality, with a median cumulative balance in non-survivors with acute kidney injury of 2.77 L compared with 0.93 L in survivors.

Interpretation in practice

The fluid balance is interpreted not as a risk score with fixed thresholds but as a trend to be set against the patient's clinical condition and treatment goals.

In septic shock and other conditions with capillary leak, a positive balance during the first few days may be unavoidable and necessary to maintain the circulation. The question is when resuscitation should end and de-resuscitation begin. A cumulative balance exceeding approximately 10 per cent of baseline body weight has been linked in observational studies to increased mortality [1, 2], but this is not a randomised, confirmed threshold and must not be applied mechanically. The assessment should integrate clinical findings, measures of perfusion, oxygen requirement and, where relevant, imaging.

Validation and performance

Since the fluid balance is an arithmetic sum and not a prediction model, measures of discrimination such as a c-statistic or AUC do not apply in the traditional sense. The central question is instead how well the measured balance reflects the patient's actual volume status and how reliable the data collected are.

In an observational study at a university clinic, the measured fluid balance was compared with daily weighings in 187 intensive care patients with 2,282 measurement points [3]. The correlation between fluid balance and change in body weight was weak (Pearson r = 0.27), even after correction for insensible losses (r = 0.27). Bland–Altman analysis showed wide limits of agreement, with a 95 per cent interval from approximately minus 4 to plus 4 kg. This means that fluid balance and body weight in practice give partly different information and that both methods should be used in parallel to detect gross errors.

A systematic review of the quality of fluid balance charting in medical and intensive care showed that in 10 of 18 studies at most 50 per cent of charts were complete [4]. Calculation errors, including the omission of intravenous drugs and drainage, were reported in 25 to 35 per cent of charts. Intervention studies with education, guidelines and visual aids improved completeness, but only 38 per cent of studies achieved at least 75 per cent complete charts after the intervention.

Limitations

The fluid balance is a sum of measured and estimated variables and depends on all fluid actually being recorded. The commonest sources of error are the omission of small volumes of intravenous drugs, incomplete documentation of drainage and vomit, and visual estimates of volumes in ungraduated containers [4]. These systematically produce a less negative balance than is real — that is, a tendency to underestimate fluid losses.

Insensible losses are an estimated, not a measured, variable. With fever, tachypnoea, a high ambient temperature or large wounds and burns, the actual losses may greatly exceed the stated 500 to 800 mL/day. This applies particularly to patients with a large body surface area or extensive skin injury.

The method cannot capture third-space losses, that is, fluid moved into the interstitium or other compartments in, for example, sepsis, pancreatitis or the postoperative state. A patient may have a negative fluid balance and at the same time be interstitially overloaded. Clinical findings of oedema and imaging may therefore be needed as a complement.

The calculator does not apply to children, in whom basal insensible losses make up a considerably larger proportion of total fluid turnover and must be calculated per kilogram of body weight. Nor does it apply to patients with large fluid losses through a stoma or fistula that do not fit the categories of the formula but must be documented under stool and other losses. Finally, the fluid balance must never be used alone as the basis for starting or stopping renal replacement therapy, but must be integrated with measures of perfusion, laboratory values and clinical assessment.

References

1. Bouchard J, Soroko SB, Chertow GM, et al. Fluid accumulation, survival and recovery of kidney function in critically ill patients with acute kidney injury. Kidney Int 2009;76(4):422–7. PMID: 19436332
2. Wang N, Jiang L, Zhu B, et al. Fluid balance and mortality in critically ill patients with acute kidney injury: a multicenter prospective epidemiological study. Crit Care 2015;19:371. PMID: 26494153
3. Mensink RSM, Paans W, Renes MH, et al. Fluid balance versus weighing: a comparison in ICU patients: a single center observational study. PLoS One 2024;19(4):e0299474. PMID: 38669249
4. Leinum LR, Krogsgaard M, Tantholdt-Hansen S, et al. Quality of fluid balance charting and interventions to improve it: a systematic review. BMJ Open Qual 2023;12(4):e002260. PMID: 38097283