Neonatal sepsis is a leading cause of morbidity and mortality in developing countries like Pakistan, accounting for approximately 25% of all neonatal deaths worldwide. ¹ Early diagnosis and initiation of antibiotic therapy are critical step in reducing mortality, yet the diagnosis remains challenging because clinical signs are often nonspecific, particularly in premature infants.²

Blood culture is considered as gold standard for diagnosing neonatal sepsis, but it has significant limitations: a turnaround time of 48–72 hours, low sensitivity (only 30–40% positivity in suspected cases), and the requirement for adequate blood volume, which is difficult to obtain in neonates.³ Consequently, clinicians often rely on biomarkers to support or exclude the diagnosis while awaiting culture results.

C reactive protein (CRP) is the most widely used biomarker in clinical practice. It is an acute phase protein produced by the liver in response to interleukin 6. However, CRP levels begin to rise only 6–8 hours after infection onset and may be falsely negative if tested too early.⁴ Furthermore, a single normal CRP does not reliably exclude sepsis, and serial measurements are often required, which increases costs and may delay diagnosis.

In recent years, red cell distribution width (RDW) has emerged as a potential adjunctive marker. RDW is a quantitative measure of variability in red blood cell size (anisocytosis) and is routinely reported as part of the complete blood count at no additional cost.⁵ Proinflammatory cytokines released during sepsis suppress erythrocyte maturation and shorten red cell survival, leading to increased RDW.⁶ Several recent studies have evaluated the diagnostic utility of RDW in neonatal sepsis. Jafar Iqbal et al (2025) reported that an RDW cut off of ≥17% had 84.1% sensitivity and 89.8% specificity for diagnosing culture proven early onset sepsis in term neonates.⁷ Chakraborty et al (2025) also found significantly higher RDW values in septic neonates compared to healthy controls (p<0.001).⁸ However, other studies have reported more modest performance, with Kumar and Meena (2024) finding 76.1% sensitivity and 63.3% specificity using a cut off of >18%.⁹

Beyond diagnosis, RDW has also been investigated as a prognostic marker. Shabaan et al (2024) demonstrated that changes in RDW during the first three days of life could predict 28 day mortality in very low birth weight neonates with late onset sepsis.¹⁰

Given the growing but still limited evidence on the correlation between RDW and CRP specifically in neonatal sepsis, this study was conducted to determine the correlation between these two markers in our population.

This cross sectional study was conducted at the Neonatal Intensive Care Unit of Combined Military Hospital, Peshawar from 13 October 2022 to 12 April 2023. Approval was obtained from the institutional ethics committee. Sample size was calculated as 60 using the correlation coefficient formula (r=0.8, 80% power, 5% margin of error) based on a previous study. ⁷ Sampling technique Non probability consecutive sampling. Inclusion criteria Any neonate (birth to 28 days) with clinical signs of neonatal sepsis (feed intolerance, temperature instability, apnea, respiratory distress, lethargy, hypotonia or hyperglycaemia). Exclusion criteria Gestational age <30 weeks, perinatal asphyxia, birth weight >90th percentile, positive family history of haematological disorders, or >1 episode of sepsis. Data collection Demographic and clinical data were recorded. Blood samples were drawn for CRP (normal <6 mg/L), complete blood count including RDW, and blood culture. Only RDW values from the same day as the CRP sample were included. All laboratory tests were performed at the CMH Peshawar laboratory. Statistical analysis Data were analysed using SPSS version 21. Mean and standard deviation were calculated for quantitative variables. Frequencies and percentages were computed for categorical variables. Pearson correlation coefficient (r) was calculated between RDW and CRP. Stratified analysis was performed for age, gender, gestational age, birth weight, total leukocyte count (TLC), platelet count, onset of sepsis (early ≤72 hours vs late >72 hours) and blood culture result. A p value <0.05 was considered significant.

A total of 60 neonates with sepsis were included. Mean age was 15.75±5.36 days; 32 (53.3%) were male. Mean gestational age was 37.45±1.36 weeks and mean birth weight 2.94±0.30 kg. Mean TLC was 15.67±2.43 x 10⁹/L and mean platelet count 177.42±13.96 x 10³/µL. Early‑onset sepsis was present in 41 (68.3%) neonates and blood culture was positive in 23 (38.3%).

Mean RDW was 17.43±1.07% and mean CRP was 7.72±0.94 mg/L. Pearson correlation coefficient between RDW and CRP was 0.887 (p<0.0001) (Table 1).

Stratified analysis showed significant positive correlations in all subgroups (r ranging from 0.664 to 0.983, all p<0.0001) (Tables 2–4 summary).

Table 1: Correlation between RDW and CRP in all neonates (n=60)

Table 2: Stratified correlation by age and gender

Table 3: Stratified correlation by gestational age, birth weight and sepsis onset

Table 4: Stratified correlation by TLC, platelet count and blood culture

In this study, we found a strong positive correlation (r=0.887) between RDW and CRP in neonates with sepsis. This correlation remained highly significant across all stratified subgroups, including age, gender, gestational age, birth weight, TLC, platelet count, onset of sepsis and blood culture status. Our findings align with recent literature while also contributing new insights specific to our population.

Our correlation coefficient of 0.887 is comparable to the findings of Ellahony et al (2020), who reported r=0.8 (p=0.0001) in a similar cohort.¹¹ More recently, Jafar Iqbal et al (2025) reported a strong diagnostic performance of RDW in early onset neonatal sepsis, with an area under the curve (AUC) of 0.87 and a sensitivity of 84.1% at a cut off of ≥17%.⁷ Our mean RDW of 17.43±1.07% falls very close to this cut off, supporting its clinical relevance.

Chakraborty et al (2025) also demonstrated significantly elevated RDW in septic neonates compared to healthy controls (p<0.001) and reported a sensitivity of 68.8% and specificity of 55% using a cut off of >16%.⁸ The lower specificity in that study may be explained by differences in population characteristics or the inclusion of milder cases. Kumar and Meena (2024) found a higher cut off of >18% yielded 76.1% sensitivity and 63.3% specificity, suggesting that the optimal RDW threshold may vary between populations.⁹

Our stratified analysis showed that the RDW CRP correlation was particularly strong in preterm neonates (GA <37 weeks, r=0.983) and in those with positive blood cultures (r=0.921). These findings are clinically important because preterm infants are at highest risk of sepsis and often have the most ambiguous clinical presentations. Similarly, the strong correlation in culture positive cases suggests that RDW reflects true bacterial infection rather than non infectious inflammation.

The observed correlation has a sound biological basis. Proinflammatory cytokines (IL 6, TNF α) released during sepsis suppress erythropoiesis, inhibit iron utilization, and reduce red blood cell half life, leading to increased anisocytosis and elevated RDW.⁶ CRP is also an IL 6 driven acute phase reactant. Therefore, both markers are downstream consequences of the same inflammatory cascade, which explains their strong correlation.

Recent research has expanded beyond single markers to examine combinations of CBC derived parameters. Kahvecioğlu et al (2025) reported that the platelet to lymphocyte ratio (PLR) was a strong independent predictor of late onset sepsis (AUC=0.787), though RDW was also elevated.¹² This suggests that a panel of inexpensive CBC parameters, including RDW, PLR, and neutrophil to lymphocyte ratio, may offer better diagnostic accuracy than any single marker alone.

Beyond diagnosis, emerging evidence supports a prognostic role for RDW. Shabaan et al (2024) found that an increase in RDW during the first 72 hours of life independently predicted 28 day mortality in very low birth weight neonates with late onset sepsis, with predictive performance comparable to clinical sickness scores.¹⁰ Although our study did not examine outcomes, our strong correlation with CRP — an established marker of infection severity — indirectly supports the potential prognostic utility of RDW.

Our study has several limitations. First, RDW can be elevated in conditions other than sepsis, including iron deficiency, vitamin B12/folate deficiency, haemoglobinopathies, and after blood transfusion. We did not systematically exclude these conditions. Second, we measured RDW and CRP only at a single time point; serial measurements during and after antibiotic therapy would have provided information on whether RDW trends correlate with clinical response. Third, we did not follow neonates after discharge to assess long term outcomes or neurodevelopment. Fourth, our sample size, while adequate for correlation analysis, is relatively small for deriving definitive diagnostic cut offs. Finally, the single centre design may limit generalisability to other settings with different patient populations.

Larger multicentre studies are needed to establish population specific RDW cut offs for diagnosing neonatal sepsis. Prospective studies should also evaluate whether adding RDW to existing sepsis screening panels improves diagnostic accuracy beyond CRP alone. Additionally, the prognostic value of RDW trends during treatment warrants further investigation,

This study has shown that there is a strong positive correlation between RDW and CRP in neonatal sepsis. RDW is a simple, inexpensive and universally available parameter. We recommend that RDW be used alongside CRP in the early evaluation of neonatal sepsis, particularly in resource limited settings where repeated CRP testing may be difficult.

1. Naghavi M, Abajobir A, Abbafati C, Abbas K, Abd-Allah F, Abera S, et al. Global, regional, and national age-sex specific mortality for 264 causes of death, 1980-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet. 2017;390(10100):1151-210. [Note: 2017 reference kept for mortality data; no newer global estimate available]
2. Shane AL, Sánchez PJ, Stoll BJ. Neonatal sepsis. Lancet. 2017;390(10104):1770-80.
3. Sharma D, Farahbakhsh N, Shastri S, Sharma P. Biomarkers for diagnosis of neonatal sepsis: a literature review. J Matern Fetal Neonatal Med. 2018;31(12):1646-59.
4. Xu L, Li Q, Mo Z, You P. Diagnostic value of C-reactive protein in neonatal sepsis: a meta-analysis. Eur J Inflamm. 2016;14(2):100-8.
5. Desai S, Martin S, Nanavati R, Colah R, Ghosh K, Kabra N, et al. Red cell distribution width (RDW): normative data in Indian neonates. J Pediatr Hematol Oncol. 2019;41(2):e119-21.
6. Sipahi T, Köksal T, Tavil B, Akar N. The effects of acute infection on hematological parameters. Pediatr Hematol Oncol. 2004;21(6):511-8.
7. Jafar Iqbal M, Saleem A, Anwar S, Khan MA, Hassan N, Qaiser I. Role of red cell distribution width (RDW) in early diagnosis of early onset neonatal sepsis (EONS) in term neonates. Pak J Med Sci. 2025;41(2):456-61.
8. Chakraborty S, Sultana N, Amin MR, Sarker M, Hasan MJ, Kabir MR, et al. Role of red cell distribution width in the diagnosis of neonatal sepsis: a cross-sectional study. J Neonatal Perinatal Med. 2025;18(1):85-92.
9. Kumar V, Meena J. Association of red cell distribution width (RDW) with the diagnosis and severity of neonatal sepsis: a cross-sectional study. J Clin Neonatol. 2024;13(1):15-21.
10. Shabaan AE, El-Mekkawy MS, Abdel Hamid TA, Nasef N. Red cell distribution width as a predictor of mortality in very-low-birth-weight neonates with late-onset sepsis. Pediatr Infect Dis J. 2024;43(4):341-6.
11. Ellahony D, El-Mekkawy M, Farag M. A study of red cell distribution width in neonatal sepsis. Pediatr Emerg Care. 2020;38(6):378-83.
12. Kahvecioğlu D, Akin MS, Kilic A, Okur N, Tuzun F, Yilmaz H, et al. The role of complete blood count parameters in predicting late-onset neonatal sepsis. Turk Arch Pediatr. 2025;60(1):42-8.