Jump to comment:

• Re: Beauty is in the eye of the beholder
  James Wynn
  Published on: 4 February 2016
• Beauty is in the eye of the beholder
  William E. King
  Published on: 6 January 2016
• Proper Use of the Heart Rate Characteristics Monitor in the NICU
  Karen Fairchild
  Published on: 23 December 2015

• Published on: 4 February 2016

  Re: Beauty is in the eye of the beholder

  • James Wynn
  • Other Contributors:
    • Jorn Hendrik Weitkamp, Jeff Reese and Ann Stark

  We appreciate the comments from Fairchild et al., and acknowledge that HRC monitoring has value in their NICU as an additional vital sign that may lead to increased provider attention. Our finding of a significantly lower correlation of HRC and proven sepsis likely stems from the difference in the definition used and highlights both a significant problem with diagnostic testing for neonatal sepsis in general (1) and the p...

  Show More

  We appreciate the comments from Fairchild et al., and acknowledge that HRC monitoring has value in their NICU as an additional vital sign that may lead to increased provider attention. Our finding of a significantly lower correlation of HRC and proven sepsis likely stems from the difference in the definition used and highlights both a significant problem with diagnostic testing for neonatal sepsis in general (1) and the practical use of HRC outside of a clinical trial. We agree that modalities that can identify infants at risk for or with disease prior to clinical presentation can provide great clinical value. However, as with any screening test for disease, it is important to know how often the test is right and how often it is wrong.

  In response to Mr. King, CEO of MPSC, manufacturer of the HeRO System, our goal was to perform an unbiased analysis of EMR data as an opportunity to share a "real-world" experience with HRC monitoring and its association with sepsis in the NICU. As we stated in our paper, this work stemmed from the fact that we were unable to identify a published report of commonly reported indices for diagnostic testing using the HeRO System including sensitivity, specificity, negative and positive predictive value, as well as area under receiver operating curve [which were also not reported in the large RCT (2)]. We would have liked to provide those metrics for our study, but were not able to do so based on sound statistical analysis. As we stated, since every baby with an elevated score > 2 was not evaluated for sepsis at each instance that the score was elevated, accurate predictive indices cannot be determined from the data available in our report (3). We acknowledged that HRC scores are monitored continuously in our NICU but not all scores were available for analysis in the EMR. Our NICU policy is that bedside staff record HRC scores at least once per shift in the EMR but also notify the provider if scores are >2.5 and record the event. Hence, the available EMR data was not entirely intermittent. In our paper, we reported not 8% of scores that were recorded but rather that 8% (9,701) of all HRC scores recorded in the EMR (127,673) over a 30-month period were > 2. The HRC score is available to integrate into clinical care as suggested, however, our data point out the limitations to accurate identification of sepsis ahead of clinical signs, and hence prevention of death from this disease. Based on our data, the actionable signal identifying sepsis by the definition we used appears hidden in a large background of elevated scores not specific to sepsis.

  James L. Wynn, J?rn Hendrik Weitkamp, Jeff Reese, Ann Stark

  1. Wynn JL, Wong HR, Shanley TP, Bizzarro MJ, Saiman L, Polin RA. Time for a neonatal-specific consensus definition for sepsis. Pediatric critical care medicine : a journal of the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies. 2014;15(6):523-8.

  2. Moorman JR, Carlo WA, Kattwinkel J, Schelonka RL, Porcelli PJ, Navarrete CT, et al. Mortality reduction by heart rate characteristic monitoring in very low birth weight neonates: a randomized trial. The Journal of pediatrics. 2011;159(6):900-6 e1.

  3. Coggins SA, Weitkamp JH, Grunwald L, Stark AR, Reese J, Walsh W, et al. Heart rate characteristic index monitoring for bloodstream infection in an NICU: a 3-year experience. Arch Dis Child Fetal Neonatal Ed. 2015.

  Conflict of Interest:

  None declared

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 6 January 2016

  Beauty is in the eye of the beholder

  • William E. King, CEO

  Regarding Coggins' Heart rate characteristic index monitoring for bloodstream infection in an NICU: a 3-year experience:

  Wirschafter[1] has labeled the CDC definition a minimum estimate of infection burden, while labeling antibiotic administration a maximum estimate. The authors refuse to provide metrics such as Specificity and NPV on the grounds that establishing health of the patients was not possible. By...

  Show More

  Regarding Coggins' Heart rate characteristic index monitoring for bloodstream infection in an NICU: a 3-year experience:

  Wirschafter[1] has labeled the CDC definition a minimum estimate of infection burden, while labeling antibiotic administration a maximum estimate. The authors refuse to provide metrics such as Specificity and NPV on the grounds that establishing health of the patients was not possible. By their own logic, meeting the CDC definition may in fact confer sickness, but the absence of meeting the definition does not confer health.

  The authors admit that only 8% of HeRO Scores were recorded, and this offered an opportunity for bias. Importantly, the recorded and unrecorded HeRO Scores were incorporated into clinical decisions during care of the subjects. The presence of elevated HeRO Scores would have prompted action on the part of caregivers that altered the presence, timing, and results of blood cultures prior to antibiotic administration[2].

  As discussed above, the authors assert that PPV, NPV Sensitivity, and Specificity cannot be calculated. Yet, the authors go on to report that "17/46 (37%) had at least one score >=2 recorded in the 48 h period prior to [BSI]." THIS IS SENSITIVITY. Further, the authors report that "BSI (at any time) was observed in just 5% of patients with HRC scores >=2..." THIS IS PPV. Back-calculated from the numbers reported:

  HeRO>=2 ever versus BSI ever:

  BSI Incidence Rate: 1.9%. Sensitivity: 87%. Specificity: 68%. PPV: 5%. NPV: 99.6%. Risk Ratio: 13.4x. ROC[3]: >.78.

  The authors characterize the 37% sensitivity versus 48hr BSI as "limited," while completely ignoring the opportunity to improve the most important outcome in a significant fraction of sick neonates. An optimist might call actionable information for 37% of proven infections in the 48 hour window prior diagnosis something quite different: "extraordinary".

  Beauty is in the eye of the beholder.

  Sincerely,

  Will King, CEO

  1 - Antibiotic use for presumed neonatally acquired infections far exceeds that for central line-associated blood stream infections: an exploratory critique. Wirtschafter DD, Padilla G, Suh O, Wan K, Trupp D, Fayard EE. J Perinatol. 2011 Aug;31(8):514-8. doi: 10.1038/jp.2011.39. Epub 2011 May 5.

  2 - Infection and Other Clinical Correlates of Abnormal Heart Rate Characteristics in Preterm Infants. Sullivan BA, Grice SM, Lake DE, Moorman JR, Fairchild KD. J Pediatr. (2014) Jan 9

  3 - ROC calculated from a single threshold (HeRO = 2.0), and the corresponding Sensitivity and Specificity (87% and 68%, respectively). Sweeping across multiple thresholds would result in a more precise calculation of ROC, which could be expected to exceed 0.80.

  Conflict of Interest:

  I am CEO of MPSC, manufacturer of the HeRO System.

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 23 December 2015

  Proper Use of the Heart Rate Characteristics Monitor in the NICU

  • Karen Fairchild, Professor of Pediatrics
  • Other Contributors:
    • David Kaufman, John Kattwinkel

  In the report "Heart rate characteristics index monitoring for bloodstream infection in an NICU: a 3-year experience", Coggins and colleagues make several observations that are important for properly using the HRC (HeRO) monitor in the NICU: 1) Continuous monitoring is more effective than intermittent. Coggins analyzed scores recorded in the medical record every 12 hours, or 8% of the hourly scores; high scores may have...

  Show More

  In the report "Heart rate characteristics index monitoring for bloodstream infection in an NICU: a 3-year experience", Coggins and colleagues make several observations that are important for properly using the HRC (HeRO) monitor in the NICU: 1) Continuous monitoring is more effective than intermittent. Coggins analyzed scores recorded in the medical record every 12 hours, or 8% of the hourly scores; high scores may have been missed. 2) Many high HRC index scores are not associated with bloodstream infection. We have found that an acute rise in the score can also be associated with urinary tract infection, necrotizing enterocolitis, clinical sepsis, acute respiratory deterioration, surgery[1], brain injury[2][3], and administration of atropine. In some cases, a rise in the score is not associated with any apparent illness or event. 3) Some infants with septicemia do not have an acute rise in the HRC index prior to diagnosis. In the Coggins report, only 37% of the 46 intermittently-sampled cases of septicemia were associated with HRC index >2. In our analysis from the continuously-sampled RCT, 79% of the 974 cases of septicemia in 700 VLBW infants, were associated with a score >2 in the day before diagnosis[4]. 4) The HRC index monitor should never replace clinical judgement. A septic-appearing baby should get antibiotics regardless of the score, and a baby with a high score but low clinical index of suspicion for infection could be closely watched without antibiotics[5]. A scenario in which a baby might benefit is the VLBW infant whose baseline HRC index is <1 for several days, followed by a rise to >2, prompting clinicians to look more closely, send laboratory tests, and detect and treat infection before the baby has obvious clinical deterioration.

  Karen Fairchild MD, David Kaufman MD, John Kattwinkel MD

  University of Virginia School of Medicine

  1) Sullivan BA, Grice SM, Lake DE, et al. Infection and other clinical correlates of abnormal heart rate characteristics in preterm infants. J Pediatr 2014;164:775-80.

  2) Vergales BD, Zanelli SA, Matsumoto JA, et al. Depressed heart rate variability is associated with abnormal EEG, MRI, and death in neonates with hypoxic ischemic encephalopathy. Am J Perinatol 2014;31:855- 62.

  3) Fairchild KD, Sinkin R a, Davalian F, et al. Abnormal heart rate characteristics are associated with abnormal neuroimaging and outcomes in extremely low birth weight infants. J Perinatol 2014;34:375-9.

  4) Fairchild KD, Schelonka RL, Kaufman D a, et al. Septicemia mortality reduction in neonates in a heart rate characteristics monitoring trial. Pediatr Res 2013;74:570-5.

  5) Griffin MP, Lake DE, O'Shea TM, et al. Heart rate characteristics and clinical signs in neonatal sepsis. Pediatr Res 2007;61:222-7.

  Conflict of Interest:

  None declared

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top