I could have cried when I clicked on the link and found that the abstract of this study ("A review of the complications of lacrimal occlusion with punctal and canalicular plugs", Rabia Bourkiza and Vickie Liee, Orbit, 31(2), 86-93, 2012) was only four sentences. Is there some mysterious medical journal etiquette requiring abstracts from literature review studies to be devoid of actual information? Oh to have free access to all the studies I'd like to read, not just the blurbs that pass as abstracts! Here it is:

Punctal and canalicular plugs are widely used for both temporary and permanent occlusion of the lacrimal puncta in dry eyes. There are many designs and materials available on the market. While their efficacy in improving dry eye symptoms is widely proven, the gamut of complications associated with these devices have never been subject to a general review, although there are numerous case series in the literature associated with one particular device. This review aims to examine the track record of a variety of plugs currently in use, to review the management of complications, and propose strategies for both the prevention of these complications and their treatment.
Knowing that "one particular device" was almost certainly SmartPlugs, this time I gritted my teeth and plunked down the fifty bucks (ouch!) for the 8 page PDF. Good though. It's truly gratifying to see in black and white in a peer-reviewed journal everything I've been yammering about for years, written by those who have actually verified from hands-on experience what I only pick up from scattered conversations. EDIT: The full study is now available as a PDF here. Thank you so much Ms. Lee!!

SO, this post is for all you skeptics who wonder why Rebecca is always getting on this weird anti SmartPlug soapbox, and why she doesn't seem to like durable intracanalicular plugs very much in general, It's not just a personal hangup. The problems are real, and they matter, and doctors, you should be sitting up and paying attention, and patients, you should be too unless you're quite sure your doctor is!

The beginning of the study describes silicone plugs of various types, SmartPlugs, and FormFit hydrogel plugs. Of SmartPlugs it says (my emphasis):

The SmartPlug is an intracanalicular plug which was originally designed to avoid the complications associated with other silicone plugs. It has the advantage of having a standard size, which makes it fit fully within the puncta, reducing the risk of irritating the ocular sur- face, and of extrusion. A study of 31 patients with the SmartPlug showed that 100% of the plugs were removed by irrigation 3 months after plug insertion. Its efficacy compared to silicone plugs was also demonstrated, with the SmartPlug showing a significantly better tear menis- cus height (mean follow-up of 11.2 weeks). However, long-term follow-up has generated a large case series of 28 patients developing canaliculitis, and a large cohort study reported a 7.23% complication rate with 4.34% developing canaliculitis.
Does anyone reading this consider 7.23% complication rate to be acceptable for a canalicular plug? And incidentally, considering there are risks of complications (some of which can be quite serious as we shall see), do you know how many doctors go through any kind of informed consent process before sticking these things in people? Not many. You wouldn't believe the calls I've had on occasion from people who had plugs put in without their knowledge, let alone consent. Anaesthetic drop... a bit of tugging sensation... done. "Um, what was that, doctor?" "Oh, I just put in some little plugs that will help keep your eyes moist." "Oh."

Moving right along. Next comes a full-page table of 14 studies in which plug complications are reported. The far right column is "Patients requiring surgery". That's our worst-case plug complications scenario, so that's what I'll focus on.

One study didn't say; seven studies (none of which was exclusively SmartPlugs) said none, two were just case reports. That leaves four case series, two of which establish incidence rates and the other two of which were simply collections of complications patients:
  • Hill et al, 2009, on SmartPlugs: 17 of 235 (7.23%) patients required surgery. The authors of this study thoughtfully pointed out that even after surgery many patients didn't get better, which is one of my biggest concerns about plug complications. (Oh, and when a study's conclusion says, "Ophthalmologists using XXX device/drug... should carefully weigh the risks and benefits..." as this one does, it's polite journalspeak for "Quit using this garbage already!".)
  • Mazow et al, 2007, on various intracanalicular plug types: 66 of 998 (6.6%) patients required surgery (I'm hoping to pin down how many per type and I will update this post when I get the info). Incidentally, in addition to the intracanalicular group of 998, there were 336 with collared punctal plugs, none of whom had complications requiring surgery.
  • SmartPlug Study Group, 2006, on SmartPlugs: 22 patients required surgery. This study noted that irrigation might help but might also make things worse.
  • White et al, 2001, on Herrick plugs: 37 patients requiring surgery, and the authors' conclusion emphasized the permanent nature of the damage incurred. Since the time of this study many more plug types have emerged.


Three out of four of the major studies about surgery resulting from plug complications involved SmartPlugs. What's wrong with this picture?

And now here's what the authors have to say about the accuracy of the complications incidence data as reported in the medical literature at large (my emphasis):

...It is likely that the complications in these studies are underestimated due to the lack of long-term follow-up. Only one study had follow-up of up to 5 years. Hill et al reported that the average time before complications developed with the SmartPlug was 3 years....
Which explains why Medennium (that's the SmartPlug manufacturer) reported NO canaliculitis or dacryocystitis cases in its 120 patient study: the 3 month follow-up period was nowhere near long enough to find out who was going to get it.

...Moreover, often it is difficult to know how many plugs were inserted, especially with the rise in number of patients undergoing laser refractive surgery where plugs are widely used in the early post-operative period. Unless meticulous records are available combined with long-term follow-up, the individual complication rates of each plug design would be impossible to evaluate.
Which brings us to one of my big pet peeves. Many doctors seem to think that if a patient had plugs put in, and some time has gone by and the patient's feeling dry again, there's no reason not to shove a fresh pair in there. But the complications data clearly bring this practice into question, as the oculoplastic surgeons who have to clean up the messes will attest to. You can't necessarily be certain the original plug is gone, and at least with the SmartPlug irrigating isn't always conclusive (or helpful) as Dr. Amy Fowler told Eyeworld a few years ago and the SmartPlug study group also noted that irrigation can actually make things worse. - Then too, if infection ensues and you need to get a plug (or plugs?) out, not knowing for certain how many there are or where won't help.

There's lots more good stuff in this study about different types of complications associated with different types of plugs, and about the truly problematic issues of plug sizing which also plays a significant role, but I'll save those for another day.

At the end of the study, the authors simply make the point I've been making for years:

Although serious plug complications are not frequent, when they do occur they may necessitate extensive surgical interventions....
To me, that's a compelling reason not to do something. If a complication is rare but "dealable", OK, you think about your risk tolerance and decide whether to go ahead. But if complications are rare but serious, and maybe not even fixable, that's an altogether different kind of assessment.

Now, though, we also have to step back and look at that statement about frequency. If you take serious plug complications as averaged across all plug types including punctal (where, with certain exceptions, for the most part the the problem is that they fall out) complications can perhaps be fairly described as "not frequent". But as we've seen here, with numbers like 7.23% or 6.6%, even the frequency may be far great for some individual plug types than most doctors and patients are aware. Finally, if the incidence is not established at all for a specific plug type - which seems to be the case for several, you may be leaping into the unknown.

Be informed before you get intracanalicular plugs.