eLetters to:

ARTICLE: Muhammad Subhani, Adriann Combs, Pamela Weber, Corina Gerontis, and Joseph D. DeCristofaro Screening Guidelines for Retinopathy of Prematurity: The Need for Revision in Extremely Low Birth Weight Infants Pediatrics 2001; 107: 656-659 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

eLetters published:

ABOUT SCREENING GUIDELINES FOR RETINOPATHY OF PREMATURITY

Jacqueline Termote, "Nicoline Schalij-Delfos, Bernard P Cats"   (16 May 2001)

Untitled

Jacqueline Termote   (22 May 2001)

response to letter

Joseph D DeCristofaro   (22 May 2001)

ABOUT SCREENING GUIDELINES FOR RETINOPATHY OF PREMATURITY 16 May 2001
Jacqueline Termote, Neonatologist, Pediatric Ophthalmologist, Neonatologist University Medical Centre Utrecht, The Netherlands, "Nicoline Schalij-Delfos, Bernard P Cats" Send letter to journal: Re: ABOUT SCREENING GUIDELINES FOR RETINOPATHY OF PREMATURITY j.u.m.termote{at}azu.nl Jacqueline Termote, et al.	JT/00537/dn ABOUT SCREENING GUIDELINES FOR RETINOPATHY OF PREMATURITY To the Editor. We enjoyed reading your article in which you advocate the revision of screening guidelines for retinopathy of prematurity (ROP) in extremely low birth weight infants. We would like to ask you some questions and present you some additional Dutch arguments. 1. One can conclude from your data, that prethreshold ROP in ELBW infants was diagnosed at a median postmenstrual age (PMA) of 32,4 weeks, with a range of 28,9 to 37,3 weeks, according to table 2. In the results however, you mention a median PMA of 32 weeks, with a range of 28 to 36 weeks. For chronological age (CA) you mention 7.5 (5.1-12.3) weeks. Do these values pertain to the 80% of infants with a PMA &#8804; 33 weeks you mention in the sentence before? Maybe this is the explanation for the difference with the value for PMA you mention in table 2. As the range in CA is smaller than the range in PMA, this suggests, to our opinion, a stronger correlation of chronological age with ROP detection. In the discussion you mention a median PMA for prethreshold ROP of 32.3 weeks, which is the value for infants without progression to threshold; we think this should be added to the text to make it easier to understand. 2. In results you further mention that 10 of 27 infants (37%) who progressed to threshold ROP were &#8804; 33 weeks PMA at diagnosis. Considering table 3, shouldn’t this be a PMA of < 33 weeks? Furthermore we are very interested in the other 15 infants who developed prethreshold ROP, especially as 7 of these (46.7%) progressed to threshold disease. 3. In the discussion you mention 10% of infants who progressed to threshold ROP to be &#8804; 33 weeks PMA, shouldn’t this be again 37% and < 33 weeks PMA? 4. To our opinion it is confusing to suggest that the most immature infants in your cohort, 23 through 25 weeks GA, had a progression rate to threshold ROP of 80%, as according to table 4, 22 of 56 patients (39.2%) with prethreshold ROP developed to threshold ROP (in discussion). However, of all the children reaching threshold disease (n=27) 80% (22/27) had a gestational age of &#8804; 25 weeks. In our department in Utrecht (a tertiary center in The Netherlands), we use a screening protocol for ROP, developed by Tan and Cats (1) using chronological age since 1984. In a review study by Cats (2) concerning the use of ICROP classification over a 13 year period (from 1984 to 1996), the detection of ROP in 307 infants was determined according to postmenstrual age and chronological age. As the slope of the curve for chronological age was found to be much steeper than that for postmenstrual age, it is clear that using chronological age for ROP screening creates a “smaller window” in which the majority of ROP cases are detected, compared to the use of postmenstrual age (*). Infants with ROP in this study had a mean gestational age of 27.7 ± 1,9 weeks and a mean birth weight of 1010 ± 293 g. The first signs of ROP in this study were detected at a mean postmenstrual age of 34.99 ± 2.53 weeks and at a mean chronological age of 7.36 ± 2.13 weeks. The smaller standard deviation in the latter also suggests a stronger correlation between ROP detection and chronological age. The same has also been described by Holmström et al (3). They found a correlation coefficient of 0,435 in chronological age vs 0,329 in postmenstrual age. (*) The figure was published in reference 2) and can be obtained by e -mailing j.u.m.termote@azu.nl. Sincerely yours, J. Termote, N.E. Schalij-Delfos, B.P. Cats Reference list: 1) Timely incidence of retinopathy of prematurity (ROP) and its consequences for the screening strategy. K.E.W.P. Tan and B.P. Cats, Am. Journal of Perinatology 1989; 6:337-40 2) Thirteen-year experience with the international classification of retinopathy of prematurity. B.P. Cats, N.E. Schalij-Delfos and K.E.W.P. Tan In: Progress in retinopathy of prematurity A. Reibaldi, M. Di Pietro, A. Scuderi and E. Molerba ed Kugler Publications B.V., Amsterdam/New York 1997; 77-81 3) Holmström G, El Azazi M, Jacobsen L, Lennerstrand G. A population based, prospective study of the development of ROP in prematurely born children in the Stockholm area of Sweden. Br. J. Ophtholmol 1993;77:417- 23.
Untitled 22 May 2001
Jacqueline Termote, Neonatologist University Medical Centre Utrecht, The Netherlands. Send letter to journal: Re: this article J.Termote{at}wkz.azu.nl Jacqueline Termote	We are very sorry, but the email-address used in the published P3R "About screening guidelines for retinopathy of prematurity" is incorrect. This should be J.Termote@azu.nl. Sorry for all the inconvenience. With kind regards, Jacqueline Termote.
response to letter 22 May 2001
Joseph D DeCristofaro, neonatology SUNY Stony Brook Send letter to journal: Re: response to letter jdecrist{at}mail.som.sunysb.edu Joseph D DeCristofaro	Drs. Terote, Schalij-Delfos, and Cats, Thank you for your interest in our paper on ROP and for giving us the Dutch experience. You were correct in pointing out a typographical error in our article where we stated in the discussion that the range of PMA for prethreshold disease was 36 weeks when it was actually up to 37 weeks as found in table 2. The semantics of gestational age seem to be confusing. In our paper when we write that the gestational age of less than 33 weeks we mean all infants who have not completed the 33rd week of pregnancy. Infants of 33.8 weeks gestational age have completed 33 weeks but have not made it to 34 weeks and we would classify them as less than or equal to 33 weeks and less than 34 weeks PMA. So that all of the statements in the article about PMA were correct as stated. The questions concerning the chronologic age for our data are as follows: CA for prethreshold disease was (mean - STD) 53.5 days +/- 10.9 days (range 36-86 days) median 53 days. The CA for threshold disease was 71.7 +/- 18.1 (41 to 117 days) median 68 days. Of those infants who progressed to threshold disease and were greater than 33 weeks PMA at the time of diagnosis of prethreshold disease (n=7) the CA was 64 +/- 7.5 days old at prethreshold disease, and 83 +/- 17.5 days old at threshold disease. They progressed to threshold an average of 18 days after the initial diagnosis of prethreshold disease. Again all infants who progressed to threshold disease were 27 weeks GA at birth or less. Of those 27 infants who did progress to threshold disease, 80% were born at less than or equal to 25 weeks GA. The point of our paper was to alert practitioners to the fact that the most immature infants are the ones at greatest risk for developing severe ROP, that it can progress rapidly in this population, that it often presents in Zone 1, and that it can occur at an early age. Thus, this population needs to be screened for ROP by around 4 weeks of age; the PMA criterion for this population should not be used. We agree with you about the usefulness of screening for ROP using CA.