Posterior blepharitis (PB) is a common, chronic, and potentially sight-threatening eyelid and ocular surface disease, characterised by inflammation and obstruction of the meibomian glands [
While PB is frequently seen, its prevalence is difficult to determine because of the lack of a standardised classification of severity [
The last International MGD Workshop (2011) summarised various grading systems used to assess MGD, focusing on meibum expressibility and quality [
This study evaluated PB both subjectively and objectively. Objective assessment of PB was performed using the Compression Of The Eyelid (COTE) grading system (Table
COTE grading system.
Grade | Nature of secretion on compression |
---|---|
1 | Clear oil |
2a | Easy egress of pus |
2b | Slow and difficult egress of pus |
3 | Thick toothpaste-like secretion (worm-like) |
4 | Complete blockage of tarsal gland; no egress of secretion visualised |
Twenty-seven patients (54 eyes) with clinically confirmed PB from a general Ophthalmology practice in Sydney, Australia, were enrolled in this prospective, consecutive case series.
The COTE test, which graded the nature and severity of expressed tarsal gland secretions (Table
Demonstration of the severity of the COTE as seen by the clinician at the slit lamp, as described in Table
The COTE test is performed at the slit lamp, using a nonpreserved, artificial tear-wetted or warm water-wetted cotton bud (Cb). COTE was performed without topical anaesthesia so that the patient would then have a sense of adequate-enough lid eversion to avoid pain due to inadvertent corneal touch with the Cb. This was emphasised during PBMP instruction, so that it could be subsequently carried out safely at home.
The central lower lid is gently everted digitally by the clinician, usually by the index finger or thumb. Firm anteroposterior pressure is applied to the posterior lid surface between that digit and the Cb at one point along the lid, effectively compressing those tarsal glands. Evaluating the central lid renders the procedure easy, rapid, and pain-free.
COTE grading was documented as grades 1, 2a, 2b, 3, or 4 (Table
Patients were then taught the PBMP technique using a plane mirror for myopes and a concave mirror, where necessary, for hypermetropes. PBMP consists of two components requiring a total of 90 seconds per session.
Patients were asked to close their eyes for 30 seconds under comfortably hot shower water and to massage simultaneously both upper lids downwards and both lower lids upwards (Figure
Massage (Part A of PBMP) of the lids. (i) Each patient is asked to close his or her lids under comfortably hot shower water and to massage, gently, both upper lids downwards and both lower lids upwards during those 30 seconds. This is done from a medial to lateral direction. (ii) The patient uses the thumb and index finger of the ipsilateral hand aiming to express the tarsal gland contents towards the lid margins.
After showering, patients were instructed to use a warm water-moistened Cb to scrub all four eyelid margins, for 15 seconds per lid (60 seconds).
Each lower and upper lid horizontal margin was divided into five equal theoretical sections. Patients were asked to scrub each of the five sections of the lid margin with the moistened Cb for three seconds, commencing medially and moving laterally.
Patients were asked to hold the Cb not too far from its tip to avoid tip instability, nor too close to reduce tip visibility; use the index or middle finger of the nondominant hand to evert and stabilise the lower lid (Figure commence lid scrubs at the most medial section of the lower lid (Figure divide the lid into theoretical fifths, allowing 3 seconds of scrubbing for each fifth. The scrubbing was carried out by moving one-fifth of the lid length laterally with each 3-second section of the scrub. By moving the everting finger for every section, it was ensured that the lid was at least at a minimum of 3 mm from the globe, and particularly the cornea, at any time point.
(a) Lid scrubs (Part B of PBMP) of the right lower lid (RLL) using a Cb on the lid margin. (i) The patient’s lid and hand have been dried and degreased by wiping with a tissue prior to lid scrubbing. (ii) In this photograph, the patient is applying lid scrubbing to the medial 1/5 of the RLL for 3 seconds. (iii) Note that the lower lid is everted adequately enough to protect the cornea, allowing precise access to its horizontal margin. (v) Note also that the Cb is held at the optimal distance from its tip to optimise patient control of the Cb tip. (b) The “Gorilla Grip” used to evert the right upper lid (RUL) away from ocular surface. (i) Note that the patient is using the Gorilla Grip on the RUL by means of the nondominant hand and middle finger. (ii) The patient is carefully inspecting the RUL with the opposite eye. (iii) The lid margin is consistently kept at least 3 mm from the globe. (v) In this photograph, the patient is carrying out PBMP on the middle 20% of the RUL. Each of the theoretical five segments of each lid takes 3 seconds.
Commencing with the medial end of the upper lid, the nondominant index, middle, or ring finger was used to achieve adequate upper lid eversion. Upper lid PBMP was also done in lid-fifths sections.
The “Gorilla Grip” facilitated this step (Figure
The principles of lower lid PBMP (Section
PBMP safety was achieved by confirming that the patient could see his or her eyelid structures at all times; the lid margin was everted at least 3 mm during PBMP; the patient was aware that the process should be pain-free, as corneal touch is obviated.
Patients received at least 15 minutes of tuition from the treating Ophthalmologist or supervising orthoptist. Patients were instructed that their competence at PBMP would then be definitively examined at one month.
Resolution of PB symptomatology was recorded on a subjective scale, expressed as a percentage of PB symptom resolution. A range of 0–100% was utilised (ranging from no improvement to complete resolution of symptoms).
Mean patient age was
COTE grading at baseline compared with post-PBMP treatment is summarised in Figure
The severity of PB was classified according to COTE criteria.
Of note, clear normal tarsal gland secretion (COTE grade 1) was demonstrated in 7.4% of tarsal glands following one month of treatment. Although not statistically significant, the novel appearance of COTE grade 1 was consistent with the trend towards objective improvement.
Reflecting the possible benefit from PBMP, patients reported an average of 77.8 ± 13.5% improvement in their symptoms at one month.
In a major survey conducted by Lemp and Nichols in 2009, 120 Ophthalmologists reported that 37% of their patients suffered from blepharitis [
Some studies [
The International MGD Workshop (2011) recommended documentation of dry eye disease, blink rate, lower lid tear meniscus height, tear osmolarity, tear film breakup time, and Schirmer’s test [
The ideal diagnostic and therapeutic approach should be straightforward enough to be rapidly performed and effective enough to achieve a convincing diagnosis for patients and clinicians.
In this study, the COTE test was used to assess MGD-related PB clinically. Similar to previous grading systems [
Although various classification systems exist, the review article of Tomlinson et al. has attempted to correlate symptoms with the number of expressible tarsal glands [
On the other hand, despite our study similarly demonstrating a lack of statistical significance in correlation (
Nevertheless, the application of COTE has the following limitations. Sampling from the central lower lid may not be representative of MGD of the whole lower lid. However, our group never evaluates the upper lid for PB using the COTE test because this would require topical anaesthesia and the conjunctival trauma would likely result in secondary corneal abrasions and patient discomfort. Whilst sampling is done at a single time point, secretory function fluctuates from time to time [ Repeated sampling over more gland locations would be helpful, but not practical, due to patient discomfort. Tarsal gland activity decreases from the nasal to temporal side of the lid. Digital pressure application applied by the Cb may vary because of a change of angle and force of Cb application even by the same examiner and variability of patient eyelid anatomy [ The COTE test may demonstrate multiple grades of severity of PB in the one lid over the application distance of the Cb, despite the fact that the lid is always assessed at its point of maximum accessibility and eversion.
Korb and Henriquez [
In a recent study, the number of expressible tarsal glands was correlated with dry eye symptoms using a custom-designed compression device, where the magnitude and duration of force were standardised [
As tarsal gland dropout occurs with increasing age due to gland atrophy rather than obstructive MGD, ageing may contribute to poor tarsal gland expressibility [
Further, numerous ophthalmological and systemic factors contribute to MGD. These include contact lens wear, giant papillary conjunctivitis, trachoma, rosacea, psoriasis, discoid lupus erythematosus, Sjögren’s syndrome, androgen deficiency, and menopause [
The COTE test has intrinsic advantages. The central 20% of the lower lid was the section that was reevaluated, providing an element of reliability. The patients were made aware that their own PBMP technique and efficacy would be mandatorily reexamined after 4 weeks of the commencement of treatment. It is possible that this degree of rigour may have resulted in enhanced compliance and subsequently improved symptomatology in patients who had failed PB treatment elsewhere.
The COTE test does have limitations, but it provides a sample of the tarsal gland characteristics and dysfunction. If PB has been diagnosed clinically and patients understand they have MGD as quantified by COTE, they can confidently be offered definitive PBMP.
In 12 years of COTE utilisation, our group has come across numerous patients who have been managed for their PB with a less rigorous technique than PBMP embodies. Simple recommendation of lid heating, followed by lid cleaning with a warm NaHCO3 solution or baby shampoo, is inadequate. The current study differs in that mandatory teaching by ophthalmological and orthoptic staff was followed by the “challenge” of an examination of each patient’s PBMP technique after a month.
The PBMP study reinforced to each patient the proper technique of PBMP. It was encouraging to note symptom improvement by 77.8%. Considering that some patients had been managed unsuccessfully previously for years, the potential benefit of our PBMP technique was clear.
Paradoxically, 9 of 54 eyes (16.7%) had worsened COTE grading, but 100% of these patients still reported an improvement in symptoms. It is possible that the residual, still-functioning meibomian glands may have improved their oil output by being rendered functional because of the effective PBMP.
Current consensus in the treatment of blepharitis [
Furthermore, this level of “lid hygiene” may not reflect the care intrinsic to our PBMP technique. The timing of each stage of “lid hygiene with warm compresses” may not be rigorously applied. While, in resistant cases, systemic oral antibiotics can be useful [
Alternative medical treatments acting by downregulating inflammatory pathways have been described. Azithromycin may suppress bacterial lipases, while oral omega-3 improves meibum lipid properties [
Newer treatments with Thermal Pulsation (TP) [
Further, IPLT causes selective photothermolysis of haemoglobin within the telangiectatic vessels of the lid [
As the PB waxes and wanes seasonally, along with the patient’s stress levels and rosacea and hormonal variations, the patient can increase the PBMP from three times a week to daily as necessary. Again, it requires only 90 seconds of the patient’s time per treatment, which is crucial considering that compliance may be an important limiting factor for effective PBMP. The most recent patient protocol for PBMP is attached (Appendix 1) in supplementary material available online at
The COTE test clearly provides a rapid, objective assessment of the expressibility of the central lower lid tarsal glands in PB. Paradoxically, the COTE test does not correlate highly with PB symptoms.
PBMP appears to be a powerful first-line treatment for PB, as patients experienced a mean symptomatic improvement of 77.8% in their PB. PBMP has been demonstrated in this study to provide a rapid, inexpensive, simple, effective, and safe method of treating PB.
Ethical approval was obtained from the Prince of Wales Hospital Human Research Ethics Committee.
All patients performing aspects of PBMP provided written consent to have their clinical photographs published in a scientific medical journal.
The authors declare that there is no existing conflicting relationship for any of them.