Is the burn centre a septic Ghetto?
This was the question raised by Raymond C. Vilain1 long before in 1977. This is very true even today, because every new arriving burn patient directly goes to burn ward where already existing infected patients, contaminated floors, beds etc. are ready to welcome this burn patients.

Sepsis remains the major cause of death in burn patients.2-4 Successful invasion of microorganisms, after resisting the defense mechanisms of body, ultimately leads to florid septicaemia. Once patient undergoes septicaemia, it can affect almost all organ systems of body inducing systemic inflammatory response syndrome followed by multiple organ dysfunction syndrome with death as an ultimate effect.

So burn patient can only be saved in early phase of septicaemia that too before the irreversible damage to various organs occurs. This requires sensitive parameter by which we can detect septicaemia in its early phase, so that early detection and institution of treatment can save the life of a burn patient. In this study, platelet count was studied as a prognostic parameter for early detection of post burn septicaemia. The present study was carried out with the following aims 1) Evaluation of platelet count and its significance in early detection of post burn septicaemia for commencement of timely vigorous treatment against it. 2) Evaluation of other laboratory parameters such as TLC, Neutrophil count and Serum Creatinine in early detection of septicaemia. 3) To study current microbial growth pattern in infected burn patients.

Material and Methods
Present study was carried out in department of surgery, Indira Gandhi Medical College and Hospital, Nagpur during the period January 2000 to June 2002.

Study Population

1. All adult (age > 18 yr.) burn cases, irrespective of sex admitted in burn ward.
2. Burns 20% to 70% of total body surface area. Burns < 20% and > 70% excluded from the study. This study has background of septicaemia in burn patients. In < 20% burns cases septicaemia is seen in very less number of patients hence excluded > 70% burns cases have very high and early mortality due to hypovolaemia and die even before developing septicaemia. So they were excluded from the study.
3. Children were excluded from the study, because of very less number of such patients available for the study making statistical inference difficult.

After admission every patient was resuscitated and scrubbed as per standard criteria and then kept on proper analgesic and antibiotic (depending on culture report) management.

Every patient was subjected to following investigations on 1st, 3rd, 7th, 14th, 21st post burn day.

1. Platelet count
   By visual method5 : 2 ml of venous blood collected in EDTA bulb 0.1 ml of blood mixed in 1.9 ml of diluent. Diluent agent used is 10 gm/L ammonium oxalate. Platelets counted on Neubauer chamber under microscope, which appear as small highly refractile particles. Platelets calculated as:
   Count/L : No of cells counted/dilution x 10⁶
   Volume counted
   Normal platelet count in adults and children is 1.5 to 3 Lac/c. mm.
2. TLC and DLC (By visual method).
3. Pus culture and sensitivity.
4. Blood culture and sensitivity.
5. Serum creatinine.
6. Blood sugar.

All the studied 66 burn patients were observed of clinical signs and symptoms of septicaemia.

Total 66 burn patients were divided in two groups, 1) Survivors and 2) Non Survivors, to analyze the laboratory parameters separately in these two groups.

Observations
Out of studied 66 burn patients, 37 patients were discharged (Survivors) and 29 patients died (Non-Survivors).

1. Study of Laboratory Parameters
Four different parameters evaluated in this study for early detection of septicaemia in burns: a) Platelet Count; b) TLC; c) Neutrophil Count; d) Serum Creatinine.

In survivors, gradual rise in platelet count was observed on subsequent post burn days. This trend was maintained till discharge of the patient. In Non Survivors, gradual decline in platelet count was observed and minimum mean platelet count was observed before death of the patient (Table 1). This pattern was not affected by the extent of burns (Table 2).

No significant variations observed in other laboratory parameters such as mean total leucocyte count; mean neutrophil per cent and mean serum creatinine in both the groups (Table 1).

2. Correlation between low platelet count and survival of burn patient
In significant number of non-survivors (62%) platelet count was low (< 1.5 Lac/c.mm) before their death while in significant number of survivors (86.5%) platelet count was normal before their discharge (Table 3).

3. Blood Culture Analysis
For evidence of septicaemia and from management point of view, blood culture was done in every patient irrespective of clinical signs and symptoms of septicaemia.

Pseudomonas (42.85%) was the predominant organism grown in blood cultures of burn patients and klebsiella (31.42%), staphylococci (17.14%) followed it (Table 4).

4. Correlation between blood culture and mortality of patients
To correlate mortality with septicaemia, incidence of positive blood culture is studied in survivors and non-survivors. It was found that blood culture was positive in significantly greater number of non-survivors (69.0%) than survivors (32.4%) (Table 5).

5. Pus culture analysis
Thus on 1st post burn day, wounds of 54.5% of the patients (36 out of 66) were sterile. Staphylococci (19.7%) and Klebsiella (14.7%) was the predominant organism responsible for burn wound infection on 1st post burn day (Table 6).
From 3rd post burn day onwards, incidence of burn wound infection ranged 88.0 to 96.1% on different post burn days. Pseudomonas was observed as predominant organism responsible for burn wound infection (28.5 to 57.5%) followed by Klebsiella (18.2 to 25%) and staphylococci (7.5 to 16.6%).

Discussion
The relationship between platelet count, WBC count and burn sepsis was shown long before in 1966 by Cohen and Gardner.6 He studied 6 gram negative septicaemic patients, irrespective of cause of septicaemia. He observed marked decrease in platelet count with occurrence of gram-negative septicaemia proved by blood culture. It was associated with leucocytosis in almost every case. He concluded that effect of bacteria or their endotoxins cause early depletion of platelet reserve leading to thrombocytopenia by bone marrow depression. However, in case of granulocytes, marked granulocytopenia does not occur due to comparative larger bone marrow reserve of granulocytes that permits destruction of WBCs without causing significant granulocytopenia.

Fig. 1 : Showing trends in mean platelet counts in survivors (S) and non-survivors (NS).	Fig. 3 : Shwoing mechanisms of thrombocytopenia in patients with burn sepsis.
Fig. 2 : Showing trends in mean serum creatinine (Cr.) level in survivors (S) and non-survivors (NS).

Housinger TA3 et al in 1993 studied relationship between platelet count, sepsis and survival of burn patient. He suggested that in septicaemia, platelet count decreases even at very low level of stimulus, which is insufficient to induce other inflammatory changes to occur. So, thrombocytopenia occurs early. Robb HJ7 in 1967 in his study of ‘dynamics of micro circulation during burn’, suggested that thrombocytopenia occurs due to huge consumption of platelets by formation of precipitates and microthrombi in arterioles and venules (Fig. 3).

Similar studies of behaviour of platelet count in burns carried out by Maduli4 et al in 1999 and by El Sonbaty2 et al in 1996. In both the studies, declining trend of platelet count was observed in non-survivors and gradual rise in platelet count was seen in survivors.

Statistical significance of difference in mean platelet counts on different post burn days in survivors and non-survivors were studied by using standard t-test. It was observed that difference in mean platelet count increases significantly on subsequent post burn days. Reason behind this is gradual rise in platelet count in case of survivors and gradual decline in platelet count in case of non-survivors with occurrence and progression of septicaemia (Table 7).

Conclusion
It can be concluded that platelet count decreases initially in all cases of burn sepsis. It gradually rises to normal in case of survivors and declines gradually in non-survivors. So serial declining platelet count can be used as prognostic indicator in burn patients for early detection of septicaemia. It helps in early institution of treatment against septicaemia resulting in favourable outcome of the patient.
Other laboratory parameters such as TLC, Neutrophil count and serum creatinine does not vary significantly with appearance and progression of septicaemia, so cannot be used as prognostic indicators of septicaemia.
Pseudomonas found to be predominantly grown organism in positive blood cultures of burn patients followed by Klebsiella and Staphylococci.

References

1. Vilain RC. Is the burn centre a septic Ghetto? Plastic and Reconstructive Surgery. Editorial 1977:793-94.
2. EL-Sonbaty MA, EL-Otiefy MA. Haematological change in severely burned patients. Annals of Burns and Fire Disasters 1996; 9 (4) : 1-4.
3. Housinger TA, Brinkerhoff C, Warden GD. The relationship between Platelet Count, Sepsis and Survival in Paediatric Burn patients. Arch Surg 1993; 128 : 65-7.
4. Maduli IC, Pati A, Pardhan NR, Panigraphy PK, Mukherjee LM. Evaluation of Burn Sepsis with Reference to Platelet Count as a Prognostic Indicator. IJS 1999; 61 (4) : 235-38.
5. Dacie and Lewis Practical Haematology : 9th edition; 19-20 and 595-97.
6. Cohen P, Gardner FH. Thrombocytopenia as a laboratory sign and complication of Gram - Negative Bacteremic Infection. Arch Intern Med 1966; 117 : 113-23.
7. Robb HJ. Dynamics of Microcirculation during a Burn. Arch Surg 1967; 94 : 776-80.