Computer Information Systems (CIS) to Assist Burn Fluid Managment

(Exhibited as a poster at ESICM 2001, Geneva, Switzerland)

MM Berger, JP Revelly, N Schai, M Lambelet, MD Bollmann, R Chiolero

Surgical Intensive Care Unit & Burns Centre, University Hospital, Lausanne, Switzerland

Rationale

During the first days after injury, patients with major burns suffer large fluid losses through their wounds, mainly oedema and evaporative and exudative losses (Figure 1): the latter last until wound closure. The insensible fluid loss has only been approximated until now in clinical settings, and roughly estimated to be 1 litre per 10% burned body surface area (BSA) from experimental data 1.

Initial fluid resuscitation is generally guided by the Parkland formula, which delivers 4 ml/kg/% BSA 2: fluid requirements are further adapted to the patients hemodynamic response. The 3rd space resulting from tissue inflammation and altered capillary permeability is extremely variable and contributes to the large fluid requirements. The fluid volumes delivered during the first 24 hours generally exceed those predicted by the Parkland formula in major burns 3. As the capillary permeability progressively reverts to normal after injury, fluid delivery is progressively decreased over the next few days. Under-resuscitation favours hypovolemic shock and consecutive organ failure. More recently, the importance of a precise fluid management in avoiding complications of over-resuscitation after major burns has been stressed: pushing fluid resuscitation too far may result in life-threatening complications 4-6.

Computerised information systems (CIS) have been introduced in clinical practice (Figure 1). They reduce nursing clerical work by capturing the physiological variables directly from the monitors and by computing many derived variables, in particular fluid balance components. We hypothetized that this calculation facility might be useful to estimate the insensible cutaneous losses in burns patients.

Aims

This prospective descriptive study aimed at assessing if computerised system could help adjust the fluid management in major burn patients, using the calculated fluid balances, and patient weight changes to determine the insensible losses.

Methods

Patients: 18 consecutive patients with major burns (> 20% body surface) admitted within the first 24 hours of injury to the ICU’s Burns, and staying more than 3 days, were enrolled (Table 1).

Fluid resuscitation: the patients were resuscitated according to the Parkland formula (4 ml/kg/% BSA) and intakes were adjusted to urine output > 0,5 ml/kg/hr and mean arterial presume > 60 mmHg. The crystalloid used during the first 24 hours was bicarbonated 0.9% NaCl 7.

Variables: Total 24hr-fluid intakes, measured fluid losses (urine, drains, aspirations), calculated fluid balances and measured daily weights were retrieved from the CIS database (MetaVision, iMDsoft, Tel-Aviv) over the first 10 days after injury. A special computer page was configured in the application to retrieve and calculate the above variables, shown in 24hr intervals (Figure 2).

Insensible losses: Last 24-hour cutaneous evaporative and exudative losses (EEL) were calculated at 12 AM as the daily
EEL = weight gain – fluid balance / 24hr

Insensible respiratory losses are included in the EEL and considered to be small and constant.

Statistics: Results in mean ± SD and linear regressions.

Results

First 24hr fluid delivery was above Parkland prediction in 17/18 patients (p < 0.01). Evolution of fluid variables is shown in Figure 3 and Table 4.

Correlation between weight changes and 24hr-fluid delivery was r2=0.365 (p<0.0001), and with fluid balances r2=0.422 (p<0.0001).

Daily EEL were correlated with burned surface (r2=0.245, p<0.0001), but not with fluid balances r2=0.093 (ns).EEL were easy to calculate in the dedicated configuration (Figure 4), and considered for next day’s free water prescription.

Conclusions

After severe burns, fluid resuscitation leads to major changes in body weight as shown by a mean increase of 25% from initial body weight. Fluid requirements frequently exceed the Parkland formula prediction.

CIS provided precise data on fluid delivery and balance, which enabled a closer estimation of the insensible losses, thereby enabling prediction of next day’s free water requirements. The fluid balances alone were an imprecise predictor of fluid gain or loss during the first days after injury due to the insensible cutaneous losses. The calculated EEL losses were generally in agreement with those from experimental designs (1 litre per 10% BSA).

Knowing the value of the previous 24 hour calculated insensible losses (EEL) was helpful information for the next day’s free water prescription. CIS may contribute to tighter fluid management, thereby improving quality of burn’s resuscitation.

References

1. Arturson G, Jakobsson O. Burns 1985; 12:1-7.

2. Baxter C. Clin Plast Surg 1974; 1:693-703.

3. Engrav L, et al. J Burn Care Rehabil 2000; 21:91-95.

4. Pruitt BJ. J Tauma 2000; 49:567-568.

5. Berger MM, Bernath M, Chiolero R. Curr Opin Anaesthesiol 2001; 14:in press.

6. Ivy M, et al. J Trauma 2000; 49:387-391.

7. Berger MM, et al. Intensive Care Med 2000; 26:1382-1385.

Table 1: Patient characteristics

Table 2: Mean fluid variables over the first 5 days in 18 patients (X±sd)

Figure 1: The CIS in the surgical ICU setting (MetaVision, iMDsoft)

 
Figure 2: Computer configured to retrieve the variables required for EEL calculation and fluid prescription in 24hr intervals

 
Figure 3: Evolution of fluid delivery, measured losses, fluid balance and weight changes over the first 10 days

Figure 4: Example of calculation in a 59 year-old patient, burned 65% BSA. These numbers were used to prescribe diuretics and adjust free water delivery for the next 24 hour periods (dextrose 5% or 0.45% NaCl)