Eye (1991) 5, 596--600

Epiphora During the First Year of Life

C. J. MACEWEN and J. D. H. YOUNG Dundee

Summary A cohort of 4,792 infants was observed in order to determine the incidence and natu­ ral history of epiphora during the first year of life. Evidence of defective lacrimal drainage was present in 964 (20%) at some time during the year. 9S�/o became symp­ tomatic during the first month of life. Spontaneous remission occurred throughout the year and 96% had resolved before the age of one. This study provides no evidence to support probing before the age of one year.

Infants with epiphora are a common problem haps the most reliable estimate of the inci­ in clinical ophthalmology. It is generally dence is 6%. This comes from a follow-up accepted that the condition is the result of a study of 200 consecutive, unselected newborn congenital abnormality of the lacrimal drain­ infants.7 age system, in the form of a membranous Information on the rate of spontaneous obstruction at the lower end of the naso-lac­ remission is also limited. The studies that are rimal duct (NLD). I In addition it is recognised available were based on small clinic popula­ that there is a high rate of spontaneous resol­ tions, referred for treatment of their epiphora ution.2•3 However, despite it's frequency, rela­ and probing was usually undertaken in a tively little is known about the incidence or number of cases before the end of the year.2,3 natural history of epiphora in young children. These series do quote high rates of sponta­ Published reports concerning the incidence neous remission, up to 100% before the age of of epiphora in infants provide widely different one,7 but the age of entry into these studies figures, ranging from 1.2% to 6% !-7 The was unclear, and a variety of 'consecutive, large variation reflectsthe different methodo­ medical' treatments were used. logies used and populations studied. For Current advice on the management of NLD example, the lowest figure of 1.2% refers to obstruction and on the timing of probing is an estimation of the incidence, based on based on unreliable data. We, therefore, con­ numbers of children with the condition sidered there was the need to determine the actually referred to hospital projected into the incidence and natural history of epiphora in a general population.4 A figure of 1.75% was large population of infants observed without derived from the incidence found in 1,538 treatment. infants attending the paediatric out-patient clinic, apparently without any follow-up.s An Methods incidence of 5% was reported by Cassady in Tayside Health Board operates a very com­ 1948, based on a general paediatric clinic prehensive computer-based paediatric sur­ population, but the ages of the children were veillance programme. Every child born in not stated, nor the reasons for referral.6 Per- Tayside is assessed at 8, 20, and 37 weeks of

Correspondence to: Dr C. 1. MacEwen, FRCS Ed, FCOphth, Department of Ophthalmology, Ninewells Hospital and Medical School, Dundee. EPIPHORA DURING THE FIRST YEAR OF LIFE 597 age. At each visit the child is examined by antibiotics or other 'medical' treatments were either an experienced health visitor or a com­ not prescribed. Parents were advised to munity paediatrician. The attendance rates simply wipe the sticky lids and lashes with are very high. cooled boiled water. This surveillance system, which has now been in successful operation for a number of Results years, was utilised as a vehicle fo,. our study. Study Population The study group comprised a cohort of all The cohort consisted of all 4,792 children children born during 1988 who were resident born between January 1, 1988 and December in Tayside at the time of their first (8 week) 31, 1988 and who were resident in Tayside at screening. With the agreement of the com­ aged 8 weeks. During the first 12 months of munity paediatricians and all local general follow-up 139 children either moved away practitioners (GPs), an additional form was from the area or were lost to follow-up and 10 included in the usual screening documents. died. Follow-up was complete on the remain­ This form required the paediatrician to ask ing 4,643 children (97% of the original the mother a series of simple standard ques­ cohort). tions, about the occurrence of any watering or To study the natural history of the condition stickiness from one or both eyes during the we evaluated each eye individually, in 73 cases baby's lifetime. Appointments were made at the condition was bilateral. From the original the eye clinic for all children who were group of 1,068 eyes which had been repork:d reported to have one or both eyes watering at as being watery or sticky, 15 were in children either of the screening times. A self-assess­ subsequently lost to follow-up. There were 16 ment form was sent to the parents and/or GPs eyes with epiphora due to epiblepharon, (all of patients who were unable to attend. had normal fluorescein disappearance tests) All children had at least two screenings by and another 14 eyes suffered isolated episodes an experienced paediatric health worker. A of conjunctivitis with no preceding of sub­ 37 week screening form was also completed if sequent history of epiphora or discharge. either or both of the childs eyes had been During the year three children with epiphora symptomatic at the 20 weeks screening. Infor­ (four eyes) had anaesthetics for the manage­ mation collected from the forms included the ment of unrelated disease. On ethical grounds ages of onset and remission of the epiphora. probing was undertaken during the same In addition a letter was sent to all GPs in anaesthetic. Therefore, the natural history of Tayside informing them of the study, and ask­ the condition was studied in the remaining ing them to refer any children with watery or 1,019 eyes. sticky eyes born during that year to the eye clinic, at the same time to avoid the use of top­ Incidence ical antibiotics except in cases with conjunc­ Of the original cohort of 4,792 children, 964 tival hyperaemia. children had evidence of defective lacrimal At the eye clinic, examination was particu­ drainage at some time during their firstyear of larly directed to the following points. Exter­ life. The incidence of this abnormality in a nal examination to confirm the presence or general population cohort was therefore 20% absence of the lacrimal puncta, lid position of all infants during the firstyear of life. By the anomaly such as epiblepharon, or cranio­ age of one year 34 children (41 eyes) were still facial anomaly. The cornea and conjunctiva symptomatic (0. 7%) were inspected. The fluorescein disappear­ ance test (FDT) was carried out on all infants Onset of Epiphora examined. This test has been shown to be very In 95% of cases the onset of epiphora or dis­ sensitive in the detection of lacrimal outflow charge occurred during the first month of life. anomalies.9.w It was used in this study to pro­ A more detailed assessment of the first two vide valuable, objective confirmation of out­ months indicates that in 834 cases the epi­ flow obstruction to complement the parents phora had been observed during their first observations. Lacrimal sac massage, topical wee.k of life, 90 in the second, 31 in the third 598 C. 1. MACEWEN AND.I. D. H. YOUNG

AGE AT ONSET OF EPIPHORA two in the 11 th and none in the 12th month. The total number of eyes with epiphora at any one time therefore reduced steadily over the course of the year (Fig. 3). There were 834

. symptomatic eyes during the first month of . � life, 608 in the sec�nd, 423 in the third, 286 in the fourth, 223 in the fifth, 164 in the sixth, 114 in the seventh, 81 in the eighth, 64 in the ninth, 53 in the tenth, 43 in the 11th and 41 in the last month of the year. The month by month resolution rate was approximately 30%, until the end of the eighth month when it declined. The overall rate of spontaneous 025 0.5 075 I 1.5 2 J 4 50 10 7 ••10 11 12 resolution by the end of the first year of life was 96%. Fig. l. Age at onset of epiphora. The firsttwo months are broken down into smaller time intervals to Probability of Spontaneous Resolution demonstrate the very early onset in most cases. In 95% The probability of spontaneous remission by of eyes epiphora presented during the firstmonth of life. the end of the firstyear of life can be evaluated from these figures (Table I). This information and a further 17 during the fourth week (Fig. may be useful when discussing the treatment 1). Another 34 started during the second possibilities with the parents of affected month of life, seven during the third month infants. For example, a child suffering from and six in the fourth month. epiphora during the third month of life has a Spontaneous Resolution 90% chance that it will have resolved sponta­ Spontaneous resolution was observed from neously by the age of one year and even as late the firstmonth through to the end of the 11 th as the ninth month, 36% will have improved month. Fig. 2 illustrates graphically the actual before the child is one year old. number that resolved during each month of life. Three hundred and sixty four eyes Infection resolved during the first month, 219 in the No child suffered from obstructed dacryocys- second, 144 in the third, 69 in the fourth, 59 in

the fifth, 50 in the sixth, 33 in the seventh, 17 NUMBER OF EYES WITHPHORA EPI EACH MONTH in'the eighth, 11 in the ninth, 10 in the tenth, 1000

SPONTANEOUS REMISSION

BOO W IL &00 0 II: W ID ... :IE o :::> "'" Z 0: .. ., ,. :> z 200

lQ II 12

" AGE (Monlhs) AGE ,Month.) Fig. 3. Eyes With Epiphora during The First Year of Fig. 2. Spontaneous remission. Each column Life. After exlusions, 4,643 infants were followed fo,' represents the number of eyes that showed resolution of the full 12 months. Each column represents that number symptoms during that month of life. or eyes watering during that month. EPIPHORA DURING THE FIRST YEAR OF LIFE 599

Table I Predicting Spontaneolls Resoilition. The table ological examination of the nasolacrimal shows the percentage of infants at each month, who on ducts of stillborn infants has previously follow-lip, resolved before aged 12 months. These demonstrated imperforate membranes in figllres can be lIsed to predict the chances of remission 70% of the cases examined.! There might also when disclissing management with parents. be a functional outflow delay as a conse­ 1 MONTH 96% quence of immature lacrimal pump function. 2 MONTHS 9>3% The early age of onset in the majority of those 3 MONTHS 90% affected strongly supports the suggestion that 4 MONTHS 86°/', tear secretion is present at birth.8 5 MONTHS 82% 6 MONTHS 75% The second important finding is the con­ 7 MONTHS 64'1'0 firmation that spontaneous resolution of epi­ 8 MONTHS 49'X, phora is very common and continues 9 MONTHS 36% throughout the firstyear of life, as indeed pre­ [() MONTHS 23'/,0 viously suggested in smaller clinical studies.2.3 11 MONTHS 5% 12 MONTHS 0% However, it has been suggested by several authors that if the condition is left untreated then this may lead to the complications of tltIs, although there were ten children with secondary infection and prolonged nasolacri­ mucocoeles. There were nine cases (18 eyes) mal duct impatency." In fact our secondary in which conjunctivitis appeared as a super­ infection rate was remarkably low with ollly added infection, secondary to defective lac­ nine episodes of secondary conjunctivitis, all rimal drainage. In all cases this settled with of which cleared quickly with the appropriate topical treatment. As previously noted, seven antibiotic treatment. The suggestion that other patients ( 14 eyes) in the cohort devel­ untreated infants might develop lacrimal flow oped conjunctivitis without any evidence of obstruction in later life seems unlikely, but associated defect in lacrimal drainage. cannot yet be refuted. Probing large numbers Discussion of infants who are, in any case, destined This report is the first large population-based to be normal must in itself carry the potential study of the incidence and natural history of risk of trauma induced lacrimal drainage epiphora in infancy and has generated a obstruction. number of significant findings. The follow-up Finally, the information contained in Table data collection was 97% complete and we are I can be usefully employed to indicate to therefore confident that our results are a true parents the likelihood of spontaneous remis­ reflectionof the incidence and natural history sion, making it easier for them to accept the in the general population. delay in surgical intervention. Firstly the incidence of 20% is much higher This study provides strong evidence to than that previously reported. The lower delay probing until at least 12 months of age. figures quoted in other studies were almost It, however, remains to be determined when certainly due to methodological problems. the rate of spontaneous remission falls to The comprehensive paediatric screening below the point when treatment should be available in Tayside has enabled us to ensure considered, thereby definingthe optimum age that all babies living in the area in 1988 were to probe. Jointly with colleagues in other included in the survey from a very early age. centres we are prospectively investigating the This has not been a feature in other studies timing of probing for all children with epi­ and the rapid improvement in the majority of phora after the age of one. cases during the firstfew months of life makes our higher figure more easily explainable. We are indebted to Sandra Murray, Barbara Hood The finding of such a high incidence of lac­ and Margaret Davidson for administrative assist­ rimal drainage dysfunction with subsequent ance. We would also like to thank Charles Guthrie remission raises the possibility that this is not and Stuart Pringle for statistical analysis. a pathological condition at all; simply a vari­ References ation of normal development in infancy. Path- I Cassady JV: Developmental anatomy of the naso- 600 C. J. MACEWEN AND J. D. H. YOUNG

lacrimal duct. Arch. Ophthalmol 1952, 47: (, Cassady JV: Dacryocystitis in infancy. Am J Oph­ 141-58. thalmol1948, 31: 773-80, 2 Petersen RA and Robb RM: The course of congeni­ 7 Guerry 0 and Kendig EL: Congenital impatcncy on tal obstruction of the naso-lacrimal duct. J Pae­ thc naso-Iacrimal duct. Arch Ophthalmol 1948, dialr Strabism1978, 15: 246-50. 39: 193-204. x Patrick RK: lacrimal secrction in full term babies and 3 Paul TO: Medical management of congenital naso­ premature babies. Trans Ophthalmol Soc UK lacrimal duct obstruction. J Paediatr 1958, 22: 1974,94: 283-9 Q. 68-70 . "Zappia RJ and Milder B: Lacrimal drainage func­ 4 Ffookes 00: Dacryocystitis in infancy. Br J Oph­ tion; 2. The fluorescein dye disappcarance test. thalmol1962, 46: 422-34 . Am J Ophthalmol 1972, 74: 160-2. 5 Stephenson S: A Preliminary communication on the to Katowitz JA and Welsh MG: Timing of initial affections of the tcar passages in ncwly born probing and irrigation in congenital nasolacrimal infants. M Press and Circ1899, 119: 103--4 . duct obstruction. Ophthalmol1987, 94: 698-705.