Objective: to relate the variables from a surveillance form for intravenous devices in high risk newborn infants with peripherally inserted central catheter related infection. Methodology: approximately 15 variables were studied, being associated with peripherally inserted central catheter related infection, this being defined by blood culture results. The variables analyzed were obtained from the surveillance forms used with intravenous devices, attached to the medical records of newborn infants weighing between 500 and 1,499 g. The statistical association was defined using the Chi-squared and Student t tests. The study was approved by the Research Ethics Committee of the Instituto Fernandes Figueira under process N. 140.703/12. Results: 63 medical records were analyzed. The infection rate observed was 25.4%. Of the variables analyzed, only three had a statistically-significant relationship with the blood culture – the use of drugs capable of inhibiting acid secretion, post-natal steroid use, and undertaking more than one invasive procedure (p-value of 0.0141, 0.0472 and 0.0277, respectively). Conclusion: the absence of significance of the variables of the form may be related to the quality of the records and to the absence of standardization. It is recommended that the teams be encouraged to adhere to the protocol and fill out the form.