4 Allergic conjunctivitis

Background

Primary eye care physicians can expect between 2 and 5% of all consultations to be related to eye conditions.1 Fifteen percent of these are caused by allergic conjunctivitis,2 which may or may not be accompanied with rhinitis (so called allergic rhinoconjunctivitis).3 Half of these cases are likely to be diagnosed with seasonal allergic conjunctivitis (SAC),4 which is a type 1, IgE mediated hypersensitivity to grass or tree pollen (the allergen). Hence, most are seen when pollens are present in the atmosphere (typically between April and August in the UK). Cases may present severely in the rare event of excessive allergen exposure. Perennial allergic conjunctivitis (PAC) has a similar immunology to SAC, except that the allergens, such as house dust mite (Dermatophagoides pteronyssinus), animal dander and moulds, are present all year round. These conditions do not affect the cornea (i.e. are non-sight threatening), and affect approximately a fifth of the population,5–7 including both genders. Racial and geographical differences in prevalence remain unclear.7 Symptoms associated with this condition, such as ocular itching and redness, often accompanied with tearing and nasal congestion, are treated by avoidance of offending antigens (although this is not always possible), topical mast cell stabilisers and topical antihistamines (with and without a vasoconstrictor).8 Systemic antihistamines can be used in association with other atopic symptoms. Ophthalmic specialists treat serious sight threatening atopic conditions such as vernal keratoconjunctivitis (VKC) and atopic keratoconjunctivitis (AKC), as these may require surgical or topical steroid intervention.

Questions

What is the comparative efficacy of topical mast cell stabilisers and topical antihistamines (with and without a decongestant) in providing symptomatic relief from allergic conjunctivitis (SAC and PAC)?

Limitations of this chapter

The efficacy of oral antihistamines in the treatment of allergic conjunctivitis, and the use of steroids and NSAIDs

for more serious conditions, such as VKC and AKC, are not considered, although these are both the subject of ongoing Cochrane reviews. In addition, the treatment of foreign body induced giant capillary conjunctivitis is not considered, as this is often complicated by the contact lens/prosthetic used.

Question

Do topical mast cell stabilisers (sodium cromoglycate, nedocromil sodium and lodoxamide) confer benefit over placebo in providing symptomatic relief from allergic conjunctivitis?

The evidence

A systematic review of all double-masked, randomised, placebo-controlled crossover and non-crossover trials comparing topical mast cell stabilisers with placebo in the management of allergic conjunctivitis, identified from the Cochrane Eyes and Vision Group Register, was performed. The authors were unaware of any previous systematic review.

Sodium cromoglycate

Eighteen double-masked studies that compared use of topical sodium cromoglycate with placebo were found. Eight studies recorded subjective symptoms whilst using treatment and placebo interventions, including ocular itching, burning, soreness and lacrimation. Five studies reported an improvement in a variety of subjective symptoms whilst using topical sodium cromoglycate preparations, and the remaining three trials found no difference in symptoms between treatment groups. Symptoms were scored and reported in different ways, with insufficient data for formal meta-analysis (this was also true of studies that contained clinician assessment of ocular signs).

Ascertainment of preferred treatment in crossover trials and overall assessment of perceived treatment benefit in non-crossover trials was more consistently reported. The characteristics of these six small trials are detailed in Table 4.19–14; note that the largest study found less difference in preference between treatment groups. Overall, a random effects estimate (due to considerable heterogeneity between

estimates; P <0·001) showed that those using topical sodium cromoglycate preparations were 17 times more likely (95% CI 4–78) to perceive benefit than those using placebo (Figure 4.1). Heterogeneity between estimates may be explained by differences in sample ages, timing of the study and/or active preparations used (see Table 4.1). The combined estimate may overestimate the beneficial effect of sodium cromoglycate, as studies that reported no differences in subjective symptoms between treatment groups did not have sufficient data for inclusion. No important side effects with the active treatment were reported, although one historic study that used phenylethanol reported stinging on installation both in treatment and placebo groups.11

Nedocromil sodium

We found five double-masked, randomised controlled trials that compared use of topical nedocromil sodium and placebo (over at least one month) for the treatment of SAC (Table 4.2).15–19 To facilitate masking, placebo drops were coloured yellow with riboflavin (0·005%) in all studies, to make them indistinguishable from the active preparation. Subjective symptoms (including itching and overall eye condition) were less in those using nedocromil sodium compared to those using placebo. The differences were statistically significant in three of these studies,15,17,18 and of borderline significance in the remaining two studies.16,19 Heterogeneity in the approaches to subjective recording of symptoms and presentation of results did not allow for a formal meta-analysis (this was also true of clinician-based assessment of treatment efficacy). Patient-perceived total and moderate effectiveness of treatment was reported in all studies (see Table 4.2). A fixed effects estimate (as differences between estimates were not statistically significant; P = 0·27) showed that those using nedocromil sodium were 1·8 times more likely (95% CI 1·3–2·6 ) to be moderately or totally controlled than those using placebo (Figure 4.2).

The patient-perceived benefit of nedocromil sodium compared to placebo appears less than the benefit of sodium cromoglycate over placebo. However, the estimate associated with nedocromil sodium is derived from studies with more participants (compare Table 4.1 with Table 4.2), is consistent between studies, and is associated with narrower confidence intervals. Apart from an unpleasant taste immediately after instillation of the active treatment, no other important side effects were reported.

Lodoxamide tromethamine

Only one randomised controlled trial of four weeks’ duration compared the use of lodoxamide tromethamine 0·1% with placebo for the treatment of allergic conjunctivitis in adults.20 Those using lodoxamide (n = 14) reported significantly less symptoms of lacrimation, burning–itching,

photophobia and lid swelling compared to those using placebo (n = 13). Fewer patients treated with lodoxamide (n = 2/14) compared to the placebo group (n = 11/13) complained of symptoms requiring additional pharmacological treatment. Three other studies using conjunctival provocation tests to a variety of allergens showed greater short-term symptomatic relief in those using lodoxamide compared to placebo. Cytological assessment in these three provocation studies found fewer inflammatory cells in the tear fluid of the lodoxamide-treated group. Alas, due to heterogeneity in methods and presentation of results, a metaanalysis of these studies could not be performed. No side effects associated with the active treatment were reported.

Comment

Although these studies consistently report improvement in symptoms of SAC in those using different topical mast cell stabilisers versus placebo, there was no evidence from randomised controlled trials to support the use of one type of mast cell stabiliser over another. Hence, treatment preferences can only be based on convenience of use (with reduced frequency of instillation for nedocromil preparations) and laboratory studies.

Question

Do topical antihistamines (levocabastine hydrochloride, azelastine hydrochloride, emedastine, antazoline sulphate) confer benefit over placebo in providing symptomatic relief from allergic conjunctivitis?

The evidence

The authors were unaware of any systematic review comparing the use of topical antihistamine preparations with placebo for the treatment of allergic conjunctivitis. In total, nine double-masked, randomised controlled trials (both crossover and non-crossover designs) on the Cochrane Vision Group database were identified; six studies compared treatment with levocabastine with placebo, one study compared azelastine hydrochloride with placebo, one study compared emedastine with placebo, and one other study from the 1970s compared antazoline phosphate with placebo (Table 4.3).21–29

Experimental25 and laboratory studies30 have shown rapid modes of action of antihistamines, especially in comparison to mast cell stabilisers. Hence, most studies used short-term conjunctival provocation tests to a variety of allergens, sometimes performed outside the pollen season, to establish the relative efficacy of topical antihistamines and placebo (see Table 4.3). A variety of symptoms and signs,

4% SCG 1 mon 8/14 15/15

Perennial

Greenbaum, RCT St Joseph’s Mean 29 yrs, Ragweed Ragweed

)Continued (

RCT patients,Torronto,Canada

12 1981

Ruggieri, RCT Istituto di Mean Feb 1983 – Active bilateral Other eye 4% SCG 4 wks 1/14 13/15

13 1987

conjunctivitis not due to infection or trauma

14 1984

VKRC, vernal keratoconjunctivitis; SCG, Sodium cromoglycate; wks, weeks; mon, months; yrs, years

including itching, redness, burning and swelling, were graded using scales ranging from 0 (none) to 3 (severe),22,24,25,29 to 0 (none) to 6 (severe)23 or visual analogue subjective scales.27 Formal meta-analysis was not possible, as most studies did not tabulate the mean scores and error associated with these scores. Often, P values associated with the difference between treatment groups were given (these are summarised in Table 4.3), but this does not allow the degree of benefit to be gauged. Despite this, most studies showed improvement in symptoms post provocation and improvement of symptoms of allergic conjunctivitis, especially itching (the hallmark symptom of allergic conjunctivitis) in those treated with antihistamines compared to those given placebo. There was no evidence from the randomised, controlled trials identified to support the use of one type of antihistamine over another.

Question

Are topical mast cell stabilisers (sodium cromoglycate, nedocromil sodium, lodoxamide) different from topical antihistamines (levocabastine hydrochloride, azelastine hydrochloride, emedastine, antazoline sulphate) in providing symptomatic relief from allergic conjunctivitis?

The evidence

The authors were not aware of any systematic review meta-analysis comparing the use of topical mast cell stabilisers with topical antihistamine preparations for the treatment of allergic conjunctivitis. Eight double-masked randomised, controlled trials comparing the use of topical mast cell stabilisers (five studies of sodium cromoglycate,31–35 one of lodoxamide36 and two of nedocromil sodium37,38) with one type of topical antihistamine (levocabastine) were identified on the Cochrane Eyes and Vision Group Register. Two were short-term trials comparing the response to conjunctival provocation to a variety of grass pollens 15 minutes after treatment with nedocromil sodium and levocabastine,37 and after 18 days of treatment with sodium cromoglycate, and four hours with levocabastine,31 in studies with 24 to 50 participants, respectively. Six trials established the longer term response to treatment with mast cell stabilisers and levocabastine in studies ranging from 14 days36 to four months in duration,32 with 3733 to 11034 study participants. Placebo drops were used to facilitate masking between treatment groups requiring different daily dosage, for example, sodium cromoglycate four times daily (qds), levocabastine twice daily (bd), ensuring an equivalent daily instillation of drops. No trials were found comparing topical

Blumenthal, Multi-centre Mean 32 yrs, Ragweed Previous seasonal Women of childbearing Nedocromil sodium 8 wks 26/71 36/69

)Continued (

Other eye disorders, Nedocromil sodium 4 wks 49/70 52/76

Moller, 4 paediatric Mean 12 yrs, Ragweed History of

UKHospital,

19 1994

† Intention to treat analysis; wks = weeks; yrs = years; RAST = radioallergosorbent test

Blumenthal, 199215

Leino, 199016

Melamed, 199417

Moller, 199418

Stockwell, 199419

Overall (95% CI)

0·1

1·81 (1·28, 2·55)

mast cell stabilisers with antihistamine and vasoconstrictor preparations.

A variety of subjective symptom scores, such as itching, tearing and burning, as well as signs (for example, redness) were graded using scales from 0 (none) to 3 (severe)31,32,34,36,37 or visual analogue scales.33,35 These were used either as separate scores or summed to give overall symptom scores. As with earlier comparisons, formal meta-analysis was not possible, as most studies did not tabulate the mean scores and error associated with these measures. Despite this, differences in scores between treatment groups were reported as not being statistically significant in the six longer term studies. A statistically significant reduction in itching and redness (P <0·05) in those treated with antihistamines was reported in the two short-term provocation studies.31,37

Mean scores were tabulated in one of these studies allowing the comparative benefit of topical antihistamine over sodium cromoglycate to be gauged.31 The benefit of antihistamine use in short studies was confirmed in interim results (at two weeks) from one of the longer studies (four

weeks).34 Patient-perceived excellent or good treatment efficacy was reported in four of the six longer term studies (these are summarised in Table 4.4). A fixed effect estimate (as there was little heterogeneity between estimates; P = 0·44) showed that those using levocabastine were 1·32 times more likely (95% CI 0·81–2·16) to perceive good treatment efficacy than those using mast cell stabilisers (Figure 4.3). However, this difference was not statistically significant. Removal of one study that compared nedocromil sodium with levocabastine (instead of sodium cromoglycate) slightly increased the perceived benefit of levocabastine over sodium cromoglycate (odds ratio 1·75, 95% CI 0·94–3·24) but this, again, was not statistically significant. The use of concomitant medications (such as systemic antihistamines, ocular and nasal medications) amongst treatment groups as a rescue medication in cases of severe symptoms, was not routinely reported and hence could not be analysed further. Despite concerns about the sedative effect with systemic use of antihistamines, there were no side effects associated with topical use.

Table 4.3 Studies comparing topical antihistamines with placebo

Abelson, RCT btn Harvard Mean 33 yrs, Not stated History of symptoms of Bacterial or viral ocular Levocabastine 4 hrs 47 eyes, Reduced itching

)Continued (

Table 4.3 (Continued)

Donshik, RCT Multi-centre Mean 36 yrs, July–Nov At least 14 years of age, Uncontrolled systemic or Levocabastine 6 wks 75, 75 Reduced itching

)Continued (

Table 4.3 (Continued)

)Continued (

Table 4.3 (Continued)

RAST, radioallergosorbent test; btn between; CPT, conjunctival provocation test; mins, minutes; hrs, hours; dys, days; wks, weeks; mon, months; yrs, years; grp, group

Conjunctivitis due to other SCG 20mg/dl (qds), 4 mon 25/34† 32/37†

Frostad, Allergologisk Median Hay fever History of

)Continued (

levocabastine (bd)

allergic conjunctivitis during the tree and grass pollen seasons, for at least 1 year, positive skin prick and RAST test to birch and thimothy grass allergens, current allergic rhinitis, aged > 15 yrs

Richard, 7 centre Mean 31 yrs, Not stated History of allergic Hypersensitivity to any Lodoxamide 14 dys 30/46 28/47

study,France yrs6–67

36 1998

Intention to treat analysis, dys, days; wks, weeks; yrs, years; SCG, sodium cromoglycate, RAST, radioallergosorbent test; qds, four times daily; bd, twice daily

Comment

Evidence from randomised control trials may show that topical antihistamines may have a quicker mode of action than mast cell stabilisers in protecting against symptoms of allergic conjunctivitis, but that there is little difference in treatment efficacy in studies of two weeks’ or longer duration. However, the relevance of acute ocular provocation studies to chronic environmental allergen exposure in the pollen season is yet to be clarified.

Summary

Despite differences in reported subjective outcomes from different studies, evidence from randomised controlled trials confirms the benefit of topical sodium cromoglycate and antihistamines over placebo preparations for the treatment of SAC. There is insufficient evidence to support the use of one class of active medication over another, hence

treatment preferences should be based on convenience of use (with reduced frequency of instillation for some preparations), risk of side effects and costs. Although we found limited evidence to suggest that topical antihistamines may have a faster mode of action than mast cell stabilisers (especially sodium cromoglycate) there was little difference in treatment efficacy beyond two weeks. Larger trials, with standardised methods and results presentation, are needed to distinguish between different topical treatments, before the increased expense of newer topical preparations can be fully justified. The use of oral antihistamines and topical steroids for the treatment of allergic conjunctivitis are the topic of ongoing Cochrane reviews.

Acknowledgements

Our thanks to Anupa Shah (Cochrane Eyes and Vision Group), who identified the references used in this chapter.

References

1.Featherstone PI, James C, Hall MS, Williams A. General practitioners’ confidence in diagnosing and managing eye conditions: a survey in south Devon. Br J Gen Pract 1992;42:21–4.

2.Manners T. Managing eye conditions in general practice. BMJ 1997;315:816–17.

3.Johansson SG, Hourihane JO, Bousquet J et al. A revised nomenclature for allergy. An EAACI position statement from the EAACI nomenclature task force. Allergy 2001;56:813–24.

4.Freissler KA, Lang GE, Lang GK. Allergic diseases of the lids, conjunctiva, and cornea. Curr Opin Ophthalmol 1997;8:25–30.

5.Dart JK, Buckley RJ, Monnickendan M, Prasad J. Perennial allergic conjunctivitis: definition, clinical characteristics and prevalence. A comparison with seasonal allergic conjunctivitis. Trans Ophthalmol Soc UK 1986;105:513–20.

6.Buckley RJ. Allergic eye disease – a clinical challenge. Clin Exp Allergy 1998;28(suppl 6):39–43.

7.Austin JB, Kaur B, Anderson HR et al. Hay fever, eczema, and wheeze: a nationwide UK study (ISAAC, international study of asthma and allergies in childhood). Arch Dis Child 1999;81:225–30.

8.Leibowitz HM. The red eye. N Engl J Med 2000;343:345–51.

9.Greenbaum J, Cockcroft D, Hargreave FE, Dolovich J. Sodium cromoglycate in ragweed-allergic conjunctivitis. J Allergy Clin Immunol 1977;59:437–9.

10.Hechanova MG. A double-blind study comparing sodium cromoglycate eye ointment with placebo in the treatment of chronic allergic conjunctivitis. Clin Trials J 1984;21:59–66.

11.Lindsay-Miller AC. Group comparative trial of 2% sodium cromoglycate (Opticrom) with placebo in the treatment of seasonal allergic conjunctivitis. Clin Allergy 1979;9:271–5.

12.Nizami RM. Treatment of ragweed allergic conjunctivitis with 2% cromolyn solution in unit doses. Ann Allergy 1981;47:5–7.

13.Ruggieri ML, Scorcia G. Double-blind group comparative trial of sodium cromoglycate eye ointment and placebo in the treatment of allergic eye diseases. Ann Allergy 1987;58:109–12.

14.van Bijsterveld OP. A double-blind crossover study comparing sodium cromoglycate eye drops with placebo in the treatment of chronic conjunctivitis. Acta Ophthalmol (Copenh) 1984;62:479–84.

15.Blumenthal M, Casale T, Dockhorn R et al. Efficacy and safety of nedocromil sodium ophthalmic solution in the treatment of seasonal allergic conjunctivitis. Am J Ophthalmol 1992;113:56-63.

16.Leino M, Carlson C, Jaanio E et al. Double-blind group comparative study of 2% nedocromil sodium eye drops with placebo eye drops in the treatment of seasonal allergic conjunctivitis. Ann Allergy 1990;64:398–402.

17.Melamed J, Schwartz RH, Hirsch SR, Cohen SH. Evaluation of nedocromil sodium 2% ophthalmic solution for the treatment of seasonal allergic conjunctivitis. Ann Allergy 1994;73:57–66.

18.Moller C, Berg IM, Berg T, Kjellman M, Stromberg L. Nedocromil sodium 2% eye drops for twice-daily treatment of seasonal allergic conjunctivitis: a Swedish multicentre placebo-controlled study in children allergic to birch pollen. Clin Exp Allergy 1994;24:884–7.

19.Stockwell A, Easty DL. Group comparative trial of 2% nedocromil sodium with placebo in the treatment of seasonal allergic conjunctivitis. Eur J Ophthalmol 1994;4:19–23.

20.Cerqueti PM, Ricca V, Tosca MA, Buscaglia S, Ciprandi G. Lodoxamide treatment of allergic conjunctivitis. Int Arch Allergy Immunol 1994;105:185–9.

21.Abelson MB, George MA, Schaefer K, Smith LM. Evaluation of the new ophthalmic antihistamine, 0·05% levocabastine, in the clinical allergen challenge model of allergic conjunctivitis. J Allergy Clin Immunol 1994;94:458–64.

22.Buscaglia S, Paolieri F, Catrullo A et al. Topical ocular levocabastine reduces ICAM-1 expression on epithelial cells both in vivo and in vitro. Clin Exp Allergy 1996;26:1188–96.

23.Donshik PC, Pearlman D, Pinnas J et al. Efficacy and safety of ketorolac tromethamine 0·5% and levocabastine 0·05%: a multicenter comparison in patients with seasonal allergic conjunctivitis. Adv Ther 2000;17:94–102.

24.Pipkorn U, Bende M, Hedner J, Hedner T. A double-blind evaluation of topical levocabastine, a new specific H1 antagonist in patients with allergic conjunctivitis. Allergy 1985;40:491–6.

25.Stokes TC, Feinberg G. Rapid onset of action of levocabastine eye-drops in histamine-induced conjunctivitis. Clin Exp Allergy 1993;23:791–4.

26.Zuber P, Pecoud A. Effect of levocabastine, a new H1 antagonist, in a conjunctival provocation test with allergens. J Allergy Clin Immunol 1988;82:590–4.

27.Horak F, Berger U, Menapace R, Schuster N. Quantification of conjunctival vascular reaction by digital imaging. J Allergy Clin Immunol 1996;98:495–500.

28.Discepola M, Deschenes J, Abelson M. Comparison of the topical ocular antiallergic efficacy of emedastine 0·05% ophthalmic solution to ketorolac 0·5% ophthalmic solution in a clinical model of allergic conjunctivitis. Acta Ophthalmol Scand Suppl 1999;228:43–46.

29.Miller J, Wolf EH. Antazoline phosphate and naphazoline hydrochloride, singly and in combination for the treatment of allergic conjunctivitis – a controlled, double-blind clinical trial. Ann Allergy 1975;35:81–6.

30.Dechant KL, Goa KL. Levocabastine. A review of its pharmacological properties and therapeutic potential as a topical antihistamine in allergic rhinitis and conjunctivitis. Drugs 1991;41:202–24.

31.Abelson MB, George MA, Smith LM. Evaluation of 0·05% levocabastine versus 4% sodium cromolyn in the allergen challenge model. Ophthalmol 1995;102:310–16.

32.Frostad AB, Olsen AK. A comparison of topical levocabastine and sodium cromoglycate in the treatment of pollen-provoked allergic conjunctivitis. Clin Exp Allergy 1993;23:406–9.

33.Odelram H, Bjorksten B, af Klercker T, Rimas M, Kjellman NI, Blychert LO. Topical levocabastine versus sodium cromoglycate in allergic conjunctivitis. Allergy 1989;44:432–6.

34.Vermeulen J, Mercer M. Comparison of the efficacy and tolerability of topical levocabastine and sodium cromoglycate in the treatment of seasonal allergic rhinoconjunctivitis in children. Pediatr Allergy Immunol 1994;5:209–13.

35.Wihl JA, Rudblad S, Kjellen H, Blychert LA. Levocabastine eye drops versus sodium cromoglycate in seasonal allergic conjunctivitis. Clin Exp Allergy 1991;21(suppl 2):37–8.

36.Richard C, Trinquand C, Bloch-Michel E. Comparison of topical 0·05% levocabastine and 0·1% lodoxamide in patients with allergic conjunctivitis. Study Group. Eur J Ophthalmol 1998;8:207–16.

37.Hammann C, Kammerer R, Gerber M, Spertini F. Comparison of effects of topical levocabastine and nedocromil sodium on the early response in a conjunctival provocation test with allergen. J Allergy Clin Immunol 1996;98:1045–50.

38.Kremer B, Tundermann A, Goldschmidt O. Onset of action, effectiveness and tolerance of levocabastine and nedocromil in topical therapy of seasonal allergic rhinoconjunctivitis. The Deutsche Rhinitis-Studiengruppe. Arzneimittelforschung 1998;48:924–30.

Background

Acute “red eye” is one of the commonest reasons for consultation with primary care physicians in the developed world, accounting for up to 5% of all clinical encounters.1–3

In the majority of cases an acute bacterial conjunctivitis is diagnosed,4–7 the pathogens most frequently responsible being Streptococcus pneumoniae, Haemophilus influenzae and Staphylococcus aureus.7–9 The condition affects both sexes, all ages, and all races.1 Generally considered to be a self-limiting disorder, antibiotics are nevertheless usually prescribed in the belief that they speed recovery, reduce the risk of developing sight-threatening complications and reduce the rate of reinfection.10–11 Guidelines on the management of conjunctivitis recommend their routine use where bacterial infection is suspected,12–14 with distinct regional preferences in the topical agent to be used.15

Question

What is the efficacy of antibiotic treatment in the management of acute bacterial conjunctivitis?

The evidence

A systematic review of the efficacy of antibiotics in the management of acute bacterial conjunctivitis is published in the Cochrane Library.16 This review includes double-blind randomised placebo controlled trials. The main outcome measures of interest were time to clinical and microbiological remission.

Three trials, studying a total of 527 patients, satisfied the review inclusion criteria. Two of the eligible trials were based in the United States,17,18 with the third recruiting patients from the United States, Mali and Morocco.19 Studies were heterogeneous with respect to age groups of patients studied, diagnostic inclusion criteria adopted by the trials, and for antibiotic treatments used (polymyxin 10 000 U/g and bacitracin 500 U/g; norfloxacin 0·3%; and ciprofloxacin 0·3%), all of which involved topical preparations. However, there was no evidence of statistical heterogeneity as revealed by the results of chi-square tests.16 The three trials included used different combinations of

outcome measures, and focused upon clinical cure, microbiological cure, or a combination of these. Clinical and microbiological outcomes were assessed “early” (days two to five post-intervention) and “late” (days six to ten post-intervention).

Meta-analysis showed acute bacterial conjunctivitis frequently to be a self-limiting condition as clinical remission

(defined as clinical cure or significant clinical improvement) occurred by days two to five in 64% (95% CI 57–71) of those treated with placebo. Treatment with topical antibiotics was associated with significantly better rates of early clinical remission (days two to five: RR 1·31, 99% CI 1·11–1·55) (Figure 5.1) and suggested that the benefit was maintained for late clinical remission (days six to ten: RR 1·27, 95% CI 1·00–1·61). Antibiotic treatment was associated with improved rates of microbiological remission, defined as pathogen eradication or reduction (days two to five: RR 1·71, 99% CI 1·32–2·21; days six to ten: RR 1·71, 99% CI 1·26–2·34) (Figure 5.2). No serious outcomes were reported in either the active or the placebo arms of these trials.

Implications for practice

This review confirms the commonly held belief that acute bacterial conjunctivitis is, indeed, usually self-limiting. The meta-analysis showed topical antibiotic treatment to be clearly associated with significantly better rates of early (days two to five) clinical and microbiological remission than is treatment with placebo, and strongly suggests that this benefit is maintained for late (days six to ten) clinical remission. No serious adverse events were noted in the 527 patients enrolled in the three antibiotic–placebo controlled trials included in this review, suggesting that important sight-threatening complications occur infrequently. But in view of the probable rarity of such adverse events the possibility of Type II errors cannot be entirely excluded.

Implications for research

The outcome measures adopted in the trials included in this review do not distinguish patient-oriented outcomes

(such as mean interval from treatment to relief of symptoms) from doctor-oriented outcomes (such as clinical and microbiological remission rates). Despite the selflimiting nature of acute bacterial conjunctivitis, and concerns regarding antibiotic safety15 and resistance,20,21 it is surprising that none of the trials attempted to determine the cost-effectiveness of topical antibiotic treatment, or assessed the impact of treatment upon re-infection rates. Such a pragmatic trial is clearly warranted and should ideally be based in primary care where the overwhelming majority of patients presenting with acute bacterial conjunctivitis are managed.22

Acknowledgements

I thank Professor Brian Hurwitz, the Cochrane Eyes and Vision Group and the Systematic Reviews Training Unit for

their assistance in preparing the original review. I am supported by the NHS R&D National Primary Care Post Doctoral Training Programme.

References

1.American Optometric Association. Optometric clinical practice guideline: care of the patient with conjunctivitis. San Francisco: AOA, 1995.

2.McCormick A, Flemming D, Charlton J. Morbidity statistics from general practice. Fourth national study 1991–2. London: HMSO, 1995.

3.Britt H, Miles DA, Bridges-Webb C, Neary S, Charles J, Traynor V. A comparison of country and metropolitan general practice. Med J Aust 1999;159(Suppl 1):S31.

4.Heggie AD. Incidence and etiology of conjunctivitis in navy recruits.

Military Medicine 1990;155:1–3.

5.Dart J. Eye disease at a community health centre. BMJ 1986;293: 1477–80.

6.Fitch CP, Rapoza PA, Owens S et al. Epidemiology and diagnosis of acute conjunctivitis at an inner-city hospital. Ophthalmology 1989;96:1215–20.

7.Gigliotti F, Williams WT, Hayden FG et al. Etiology of acute conjunctivitis in children. J Pediatr 1981;98:531–6.

8.Mahajan VM. Acute bacterial infections of the eye: their aetiology and treatment. Br J Ophthalmol 1983;67:191–4.

9.Perkins RE, Kundsin RB, Pratt MV, Abrahamsen I, Leibowitz HM. Bacteriology of normal and infected conjunctiva. J Clin Microbiol 1975;1:147–9.

10.McDonnell PJ. How do general practitioners manage eye disease in the community? Br J Ophthalmol 1988;72:733–6.

11.Morrow GL, Abbott RL. Conjunctivitis. Am Fam Physician 1998;57:735–46.

12.Donahue SP, Khoury JM, Kowalski RP. Common ocular infections: a prescriber’s guide. Drugs 1996;52:528–9.

13.Canadian Ophthalmological Society. Eye conditions, disorders and treatments. http://www.eyesite.ca/info/04ae-condconj.html (accessed 25 Feb 2000).

14.British Medical Association and Royal Pharmaceutical Society of Great Britain. British National Formulary. London: BMA, 1999, p 38.

15.Doona M, Walsh JB. Use of chloramphenicol as topical eye medication: time to cry halt? BMJ 1995;310:1217–18.

16.Sheikh A, Hurwitz B, Cave J. Antibiotics for acute bacterial conjunctivitis (Cochrane Review). In: Cochrance Collaboration: Cochrane Library. Issue 2. Oxford: Update Software, 2003.

17.Gigliotti H, Hendley JO, Morgan J, Michaels R, Dickens M, Lohr J. Efficacy of topical antibiotic therapy in acute conjunctivitis in children. J Pediatr 1984;104:623–6.

18.Leibowitz HM. Antibacterial effectiveness of ciprofloxacin 0·3% ophthalmic solution in the treatment of bacterial conjunctivitis. Am J Ophthamol 1991;112:29s–33s.

19.Miller IM, Wittreich J, Vogel R. The safety and efficacy of topical norfloxacin compared with placebo in the treatment of acute, bacterial conjunctivitis. Eur J Ophthalmol 1992;2:58–66.

20.Chaudhry NA, Flynn HW, Murray TG, Tabandeh H, Mello MO, Miller D. Emerging ciprofloxacin-resistant Pseudomonas aeurginosa.

Am J Ophthalmol 1999;128:509–10.

21.Goldstein MH, Kowlaski RP, Gordon YJ. Emerging fluoroquinolone resistance in bacterial keratitis: a 5-year review. Ophthalmology 1999;106:1313–18.

22.Sheikh A, Hurwitz B. Topical antibiotics for acute bacterial conjunctivitis: a systematic review. Br J Gen Pract 2001;51:475–7.