There are several hematological findings for early diagnosis of neonatal sepsis before availability of the blood culture results. Thrombocytopenia is one of these (
3). Thrombocytopenia occurs in half of the cases with proven sepsis (
10) and is used as a marker for screening of neonatal sepsis. It has been reported to be more common in association with some sepsis causing germs. This will be helpful to choose appropriate empirical therapy soon before availability of blood culture results. Ahmed et al. (
11) reported to have seen thrombocytopenia more common in gram negative sepsis and that it had more prolonged duration in these cases than in Gram-positive sepsis. Some authors have even seen severe thrombocytopenia with Gram-negative sepsis (
4). In addition, it correlated with the outcome of the patients especially in VLBW newborns (
12). In a study Ahmad et al. (
12) showed that thrombocytopenia was more common in fungal sepsis and Bhat et al. (
13) reported high incidence of thrombocytopenia with Gram-negative and fungal infections. In addition prolonged thrombocytopenia, very low platelet levels, and need to platelet transfusion were poor prognostic factors. In a similar study by Alshorman Abdullah et al. thrombocytopenia was detected in 25% of infants with sepsis, but significant differences were not observed in Gram-positive and negative infections (
14). All studies confirmed thrombocytopenia to be common in septic neonates, but can thrombocytopenia determine different types of neonatal sepsis germs, too?
Other platelet findings in diagnosis and predicting of sepsis are PDW and MPV (
15). Elevated MPV (
16) and PDW may be used as a marker for screening of sepsis. Choudhary et al. (
17) reported that high MPV, high PDW and thrombocytopenia are more common in late onset than early onset sepsis. Although, Shaaban and Safwat (
18) found MPV as a marker of early onset sepsis. Catal et al. (
1) estimated that elevated MPV can be a diagnostic factor along with other findings and can even be utilized for following the course of the sepsis in neonates but, in line with our study, they didn’t find any correlation between elevated MPV and different kinds of sepsis germs. Akarsu et al. (
19) also didn’t find any correlation between elevated MPV and PDW and different germs of sepsis, although we found that elevated PDW is more common in Gram-negative infections (P value = 0.026). Some authorss reported that septic patients with PDW levels greater than 18% have worse outcome (
20). Ahmad and Waheed (
3) did not accept this and introduced thrombocytopenia instead as the predictive factor in neonatal sepsis and found higher mortality rates in thrombocytopenic neonates.
5.1. Conclusions
Evaluating platelet indices in neonates with different sepsis germs we found that, although thrombocytosis may be a laboratory finding in septic neonates, it is not common in early onset infections (P value = 0.028). In addition, thrombocytopenia, elevated MPV or PDW can be laboratory findings for diagnosis of sepsis but are not specific for special types (namely early, late sepsis, nosocomial infection) of infection. Thrombocytosis, thrombocytopenia and elevated MPV are seen in Gram-positive, Gram-negative and fungal infections and they are not specific findings in determining different types of sepsis germs but PDW can be helpful because it was observed more common in Gram-negative infections in the present study (P value = 0.026), although the mean value of PDW in Gram-negative infections was not more than that in cases caused by other germs.
So, note that:
• In a young neonate with thrombocytosis sepsis is not the first thing to be considered because thrombocytosis is not common in early onset sepsis.
• In septic neonates pay attention to platelet indices, if PDW is over 17%, Gram-negative infections are more probable than Gram-positive and fungal infections. This is important for starting empiric therapy.
We admit that we had limitations in our study. The study was carried out in out-born hospital patients, so we had only few neonates with early onset sepsis. In addition, the number of studied neonates was limited. So, multicenter studies with larger sample size are necessary to come to more reliable conclusions.