| HOME | HELP | CONTACT US | SUBSCRIPTIONS | CME | ARCHIVE | SEARCH |
|
|
|
||||||||
|
||||||||||
E-Letters to:
|
|
![[Read E-Letter]](/icons/misc/sectionDOWN.gif){kind=icon}
CRP: Role in Monitoring Response to Treatment is Open to Question
CRP: Response to Dr. Piecuch
C-reactive protein and waterbirths?
Response to Dr Costakos
|
|
|||
|
Steve Piecuch, MD, MPH, Neonatologist Department of Pediatrics, SUNY-Downstate Medical Center, Brooklyn, NY
Send letter to journal:
stevepiecuch{at}aol.com Steve Piecuch, MD, MPH
|
I found the article by Weitkamp and Aschner on the use of CRP in the evaluation of neonates with possible infection to be informative and useful to my own clinical practice. They make a convincing case that CRP provides information in addition to that provided by history, physical examination, CBC and blood culture in assessing the infant at increased risk of infection who may benefit from empiric antibiotic therapy. Since CRP can assist in identifying those infants more likely to be infected despite negative cultures, it makes sense to consider CRP when deciding whether to continue antibiotics beyond 48-72 hours if cultures are negative, particularly in cases of possible early-onset infection where the mother received intrapartum antibiotics. Weitkamp and Aschner propose a protocol which includes doing an end- of-treatment CRP prior to discontinuing antibiotics in some cases. In fairness, these authors advocate a relatively limited role for CRP in assessing response to treatment and their protocol is quite reasonable. They make it clear that decision-making cannot be based upon CRP or any other laboratory test in isolation and call for additional studies in this area. However, I would like to go beyond the recommendations that they make and raise the more general question of whether the routine use of CRP to monitor response to treatment is appropriate. It has been my experience that some clinicians not infrequently utilize CRP for this purpose. I would argue that adequate evidence to support such a role does not exist. Consider a term neonate with negative blood cultures who has been treated for 7 days for “clinical sepsis” and who has been doing well for 5 days. Many neonatologists would discontinue antibiotics without performing any additional tests. What evidence supports the usefulness of an end-of- treatment CRP in such an infant? Do we know what percentage of such infants have an elevated CRP at the end of treatment? Of those that do have an elevated CRP, is there evidence that continued antibiotic therapy is necessary? Can we rely upon CRP to identify the occasional infant such as this who does require more prolonged treatment? Do we know what percentage of those with persistent infection will have a normal CRP? What about the infant with negative cultures who has been treated for 7 days and is doing well but who never had a spinal tap? Can we rely upon a normal CRP to exclude partially treated meningitis which might require continued antibiotic therapy? Consider the patient with group B streptococcal meningitis who was initially very ill and whose CRP was significantly elevated. If the patient is responding as expected and the CRP is coming down, is it necessary for the CRP to be normal before stopping antibiotics after 14 days of treatment? Do we have evidence-based answers to these questions? This is not to say that CRP may not have a role in monitoring patients with infection. We can all probably agree that in most cases the CRP goes up in the presence of infection and comes down as the infection is treated. A CRP approaching the normal range may be a reasonable point at which to switch from parenteral to oral antibiotics in a patient with osteomyelitis or septic arthritis. Normalization of CRP may be reassuring in a patient with a complicated infection, and a new increase in CRP may be an indication that additional investigations or a change in antibiotic regimen is indicated. My point is that there is insufficient evidence to support routinely monitoring CRP during treatment or doing an end-of- treatment CRP in patients who are responding as expected according to generally accepted clinical and laboratory parameters. The amount of weight that should be given to a normal CRP in a clinically ambiguous situation is also open to question. While CRP clearly has a role in screening for infection, its practical role in monitoring response to treatment remains undefined. Conflict of Interest:None declared |
|||
|
|
|||
|
Judy L Aschner, neonatologist Vanderbilt Children's Hospital, Jörn-Hendrik Weitkamp
Send letter to journal:
judy.aschner{at}vanderbilt.edu Judy L Aschner, et al.
|
In Reply – We thank Dr. Piecuch for his thoughtful commentary on our recent article, "Diagnostic Use of CRP in Assessment of Neonatal Sepsis" and were gratified that he found our contribution to this literature useful to his clinical practice. In his comments, he raises the question of whether the routine use of CRP to monitor response to treatment is appropriate and, in illustration, lists several clinical scenarios. He rationally argues that in such cases adequate evidence to support a diagnostic role of CRP does not exist. Indeed, we attempted to emphasize that point, advocating a limited role for CRP in this assessment until data validating its usefulness are available. The first case that Dr. Piecuch describes is a term neonate with negative blood cultures who has been treated for 7 days for “clinical sepsis” and who has been doing well for 5 days. This is a typical case of a neonate with presumed sepsis and negative cultures. According to our protocol, we would have guided therapy based on serial CRP values and discontinued antimicrobial therapy after 48 hours in a stable patient with 2 negative CRP values and negative cultures, provided the patient was not neutropenic. In the case of an initial or 48 hour positive CRP value, we would perform a full sepsis-workup, including blood-, CSF-, and (if the patient is older than 72 hours) urine-cultures. The results of these cultures in combination with the history and physical examination would determine antimicrobial choices and length of therapy. Similar to the repeat lumbar puncture in patients treated for meningitis that is recommended by most experts, we believe that a negative CRP after 7 days of therapy is reassuring that the infectious process has been treated successfully. A continued positive or rising CRP suggests insufficient response to therapy. In this case further diagnostic procedures may be warranted. Dr. Piecuch raises important and currently unanswered questions: Of those infants that have an elevated CRP, is there evidence that continued antibiotic therapy is necessary? Can we rely upon CRP to identify the occasional infant who requires more prolonged treatment? Do we know what percentage of those with persistent infection will have a normal CRP? Can we rely upon a normal CRP to exclude partially treated meningitis which might require continued antibiotic therapy? As we pointed out in our review, despite the widespread use of CRP, prospective data or data from randomized controlled trials are very limited. Similarly, decisions about the length of therapy for culture-negative and even many culture-positive infections in neonates are predominantly based on expert consensus. We do acknowledge the lack of data, but would reiterate that decision-making in the absence of CRP data is no more evidenced-based than with it. Serial CRP values are simply one more tool in the hands of the experienced clinician to guide length of antimicrobial therapy. In our article, we shared with the readers a CRP protocol currently in use at Vanderbilt Children’s Hospital. Since implementation of this CRP protocol via our Computerized Physician Order Entry (CPOE) system, we have collected data prospectively on compliance with the protocol and outcomes for the patients with regard to subsequent culture results, clinical course and antimicrobial use. We look forward to providing more objective evidence of the value or lack of value of CRP in monitoring the response to treatment following completion of our prospective (non-randomized) study. Jörn-Hendrik Weitkamp, MD Judy Aschner, MD Conflict of Interest:None declared |
|||
|
|
|||
|
Dennis T Costakos, neonatologist Mayo health System
Send letter to journal:
costakos.dennis{at}mayo.edu Dennis T Costakos
|
Dear Drs Jörn-Hendrik Weitkamp and Judy Aschner, Thank you for your well written article concerning C-reactive protein published in the november 2005 Neoreviews. I wonder if you can comment on C-reactive protein (CRP)kinetics in the setting of second stage occurring in a water bath? (waterbirth).(Particularly if the baby looks well.)When I review the literature, I do not find a single paper looking at this, and in some parts of world, including one town in Wisconsin, a high percentage of babies are born in water. always , Dennis T. Costakos, MD, FAAP Conflict of Interest:None declared |
|||
|
|
|||
|
Judy L Aschner, neonatologist Vanderbilt Childrens Hospital, Jörn-Hendrik Weitkamp
Send letter to journal:
judy.aschner{at}vanderbilt.edu Judy L Aschner, et al.
|
We thank Dr. Costakos for his comments. Dr. Costakos raises the interesting question about literature on C reactive protein (CRP) kinetics in the setting of waterbirths. This is a valid point in an environment that performs waterbirths on a regular basis, because the mode of delivery must be taken into account when using CRP for the assessment of early- onset sepsis. Waterbirths are increasingly being offered as an option (particularly in Europe), although concerns about associated infection risks have been expressed and special infection control policies for waterbirths are recommended (J Hosp Infect. 41 (1999):155-157). As Dr. Costakos points out, data comparing CRP values between waterbirths and regular births are very limited. We found an observational study by Geissbuehler et al. from Switzerland reporting a nine year experience comparing "landbirth" and "waterbirth". The authors found that waterbirths were less associated with "injuries" compared to landbirths and that woman undergoing waterbirths were less likely to present maternal risk factors including chorioamnionitis. No case of water aspiration or other perinatal complication of the mother or child which might be water-related was reported. When all types of infections were considered, significantly fewer neonatal infections were seen among waterbirths (J. Perinat. Med. 32 (2004):308-314). Our personal experience with waterbirths is very limited but based on the literature cited above, we would expect similar CRP kinetics as with uncomplicated vaginal births. CRP concentrations in newborns peak between 24 to 48 hours after birth and can exceed the cutoff of 10 mg/L, especially after vaginal delivery that involves instrumentation. Normal range CRP values after waterbirths versus regular vaginal birth versus assisted delivery versus C-section are difficult to define, because CRP values in these settings are influenced by several confounding factors including the interval between rupture of membranes and delivery, length of active labor, and neonatal factors, such as gestational age and birth weight. However, we would encourage Dr. Costakos to add to the literature and prospectively compare CRP values in infants born by waterbirth versus landbirth. Jörn-Hendrik Weitkamp & Judy Aschner Conflict of Interest:None declared |
|||
| HOME | HELP | CONTACT US | SUBSCRIPTIONS | CME | ARCHIVE | SEARCH |
E-Letters: