Introduction

Epiphora impacts on quality of life in adult patients by causing blurred vision, spattered glasses, and sore skin.1, 2 Tearing is socially embarrassing as it mimics the appearance of persistent crying. An important cause of epiphora is narrowing or occlusion of the nasolacrimal duct.3 Nasolacrimal duct blockage is circumvented by surgically creating an anastomosis between the lacrimal sac and nasal cavity above the site of occlusion, by either external dacryocystorhinostomy (DCR) or endoscopic endonasal dacryocystorhinostomy (EE-DCR). The external approach has remained the mainstay of treatment for over a century.4 With the advent of fibreoptic endoscopes utilising the principles of total internal reflection and the simultaneous use of a light pipe within the lacrimal sac to guide placement of the osteotomy, the endoscopic endonasal approach to DCR gained popularity from the 1990s.5 Using a laser to create the rhinostomy is outdated because of high rates of fibrosis secondary to charring, with mechanical approaches using a diamond bur showing greater successes.6, 7

We offer all patients endoscopic endonasal surgery, with its advantages of no scar, shorter postoperative recovery, greater heamostasis, and shorter operating time. In this study, we aimed to measure the impact of EE-DCR on patients’ quality of life, using the Glasgow Benefit Inventory (GBI).

Robinson et al8 developed the GBI, a post-interventional questionnaire to be used in otolaryngological (ORL) procedures. It is a validated questionnaire that is both sensitive to the change in health status due to a surgical procedure and patient orientated. It is comprised of 18 questions, each of which is based on a five-point Likert scale.8 The questions are specifically tailored to measure a change in health status, defined as the general perception of well-being (12 questions). Social and physical health parameters are also assessed, with three questions each. The total GBI scores from −100 (maximal negative benefit), through zero (no change) to +100 (maximal positive benefit in health status).8 Any positive score represents patient satisfaction with the intervention. Robinson aimed to control response bias by having half the answers range from large improvement to large deterioration and the remaining half conversely. We applied the GBI to collect data for EE-DCR, as quality of life is a significant contributor to overall success of the procedure.

Materials and methods

A postal questionnaire was sent out retrospectively, at least 6 months post EE-DCR. The study was carried out in accordance with ethical guidelines of the Declaration of Helsinki with institutional ethics approval. Patients were listed for surgery based on clinical assessment (including syringing and probing), plus/minus radiological diagnosis with lacrimal scintigraphy. Those with nasolacrimal duct obstruction (NLDO) were offered EE-DCR surgery.

Endoscopic endonasal DCR was performed under general anaesthesia. All patients had the nasal mucosa decongested with Moffats solution, combination of 1 ml adrenaline 1 in 1000, 2 ml of 10% cocaine, and 2 ml sodium bicarbonate. The lateral nasal wall was infiltrated with 1 ml of 2% xylocaine (with 1 in 200 000 adrenaline). The lacrimal crest was identified using a rigid 0° and/or 30° Hopkins endoscope. The nasal mucoperiosteoum over the frontal process of the maxilla and lacrimal bone was elevated. A Medtronic diamond bur was used to create the bony rhinostomy. A fibreoptic light pipe was inserted into the lacrimal sac, enabling visualisation and subsequent vertical incision of the sac and flaps with a keratome. An oscillating blade was used as needed to fashion mucosa. O’Donoghue tubes are inserted and knotted.

Theatre records and operation notes were reviewed, and demographic data collected. Patients were eligible for our study if they had a NLDO and were aged 18 and over. Exclusion criteria included patients undergoing external or laser procedures, revision procedures for failed primary surgical procedures, pregnancy and secondary acquired NLDO (such as from sarcoid or Wegener’s granulomatosis), and those with obstruction at the canaliculi or common canaliculus.

Each patient was sent the GBI questionnaire, enclosed with an information leaflet, consent form, and a stamped addressed envelope (see Appendix). We gave patients a 3-month period to respond, after which they were called twice, once during the day and once in the evening. We entered the data into an excel database and statistical analysis was performed.

Results

The records of 282 consecutive patients who underwent EE-DCR surgery were analysed. We sent out 250 surveys, as 32 addresses were not available. Out of the responses, 14 were returned anonymously and hence we were unable to add demographic data.

We received 97 posted responses, 3 months from sending out the questionnaires, representing an overall response rate of 38.7%. Three were incomplete and hence were not included in the analysis. We than rang all non-responders to improve the compliance rate, and overall were able to fill 16 further questionnaires over the phone, giving a total of 110 responders.

The mean age of patients, at surgery, was 62 years (95% CI: 59.36–65.08), with a range of 18–94. Overall 63% were female, 37% were male and 56% were British. From the 110 questionnaires analysed, the mean total score from the GBI was +15.04 (95% CI: 9.74–20.35). The general subscale score was +22.16 (95% CI: 15.23–29.09). Social support scale resulted in a mean of +4.67 (95% CI: 0.93–8.42) and physical health scored −4.47 (95% CI: −10.25 to 1.32). Figure 1 shows box-plots of both total and subscale scores, when individual ratings are categorised as either plus or minus scores.

Figure 1
figure 1

(a) Box plot of total scores. Total GBI scores, broken down into separate plus and minus scores. (b) Box plot of subscale scores. General GBI scores, broken down into separate plus and minus scores.

Full size image

Table 1 shows a breakdown of the operating surgeon in all 110 procedures. A t-test was performed for rank of operator and showed that consultant only (n=67) had a mean total score of 13.18 (95% CI: 7.14–19.21). The other group included any patients who had a trainee operating on them (even if that was together with a consultant, n=26) and had a mean of 20.82 (95% CI: 9.12–32.54). This difference was not statistically significant (P=0.2199), similar to functional successes noted by Fayers et al.2

Table 1 Representation of all procedures performed by various operators
Full size table

A t-test for total score by patient age category was performed, split at the median of 63.5. The younger group’s mean was 19.04 (95% CI: 11.35–27.74), compared with 11.04 (95% CI: 3.61–18.47) in the older group.

From undergoing surgery to receiving the questionnaire, there was a mean time of 47.01 months (with a range of 6–133).

Discussion

In this study, we have shown by using a validated tool that quality of life improves with EE-DCR. Modern medicine has moved full circle from the paternalistic approach practiced by previous generations. An important requirement of contemporary medicine is the patient experience: with particular emphasis on measuring the effect of an intervention on patient benefit. Vast improvements occurred in this area when Robinson et al8 devised the GBI, a validated questionnaire enabling assessment of the patients’ health status post otolaryngeal or ophthalmic intervention.

We sent the questionnaires by mail, without seeing the patients. We aimed to avoid clinical bias from objective measures and use patient experience exclusively to assess the success of the procedure. To further highlight the importance of using the patients’ experience as the primary outcome measure, Tarbet and Custer9 found that 62% of all patients with patent DCR’s to irrigation still had persistent epiphora clinically. Furthermore, Delaney and Khooshabeh10 described only 38% of patients with patent DCRs clinically classed themselves as completely asymptomatic through questionnaire. It should be kept in mind that a significant number of these patients may represent epiphora of multifactorial cause.

We chose outcomes of at least 6 months postoperatively to assess success.2, 7, 11 Our response rate of 38.7% was markedly lower than other postal questionnaire studies for OLR procedures12, 13 and most likely reflects our population of an inner city multi-cultural mix of patients, with 44% of the study population non-British. We found that the EE-DCR gave patients improvement in areas of general perception of well-being, including the social and psychological components contributing to health. The predominantly positive scores demonstrate that patients perceive EE-DCR as a beneficial procedure, with improvement in parameters including mucoid discharge, blurry vision, and soreness peri-orbitally from persistent wiping of the skin. Unlike the other three parameters, the mean physical health score was −4.47. We were unable to fully account for this negative finding, but analysis of the GBI questionnaire shows that the questions enquiring about physical well-being are quite generic and relate to systemic health, with little correlation to the original symptoms (refer to questions 8, 12, and 16). Overall, as well as a lack of scarring, shorter operating time and minimal blood loss, EE-DCR now also receives positive feedback from validated patient assessments, giving credence to offering it as a first-line management of NLDO.

The ENT literature has published the use of questionnaires to assess patient symptoms post EE-DCR procedures, and authors have shown patient subjective improvement of symptoms. For instance Agarwal14 followed up 300 patients a year after EE-DCR and noted 94% were symptom free after the primary intervention. Validated quality of care data were not collected. Karim et al15 found 84% in the endoscopic cohort were symptom free at least 4 months postoperatively. Zenk et al16 gained a greater perspective into long-term symptom improvement, after performing a retrospective analysis of 165 patients who underwent EE-DCR. Total or partial resolution of symptoms, as described by patients, occurred in 81.8% (total resolution in 67.9%).

The GBI has been used in a wide variety of ORL procedures, including rhinoplasty,17 acoustic neuroma surgery,18 endoscopic sinus surgery,12 and EE-DCR.1 Ho et al19 used the GBI questionnaire in a prospective study of 55 patients undergoing EE-DCR. The patients were reviewed at 6 months by a consultant ophthalmologist and graded as a success if the symptoms were either cured or better. Surgery was successful in 78% of patients; the mean total GBI scores for successful and failure groups were +34 and −19, respectively (comparable with figures from our study of +23 and −24, respectively). This study demonstrated that the GBI is a sensitive measure, as it can differentiate between success and failure. Most recently, Hii et al20 have prospectively evaluated adults treated with external or EE-DCR, analysing patient satisfaction with the GBI, economic cost, and surgical success. The study differed from ours as it was prospective, the authors aimed to compare the two procedures directly, and objective measures of success were also observed. Thirty-seven patients underwent external DCR and 40 were in the EE-DCR group. The questionnaire was posted 6 weeks postoperatively and the mean GBI scores were +16.1 and +24.1, respectively. The difference of 8 between the scores did not reach statistical significance,20 suggesting that although both operations produce positive post-interventional change in health status, the difference between both is negligible. As anatomical and functional success at 3-month follow-up in both groups were similar (91.7 and 92.1% in the external and EE-DCR groups respectively) and the cost difference did not reach statistical difference,20 one can conclude that patient choice and surgeon preference key in deciding what approach is adopted. We feel the advantages of no scar, shorter postoperative recovery, greater heamostasis, and shorter operating time advocate the use of EE-DCR as front-line surgery, substantiated with the results from the Hii et al20 study showing no difference between the surgeries in three parameters.

In our study, we found that younger patients had statistically significant improved general perception of well-being compared with older patients. This is a consistent finding in the literature, for instance, Tripathi et al21 showed a statistical correlation between complete resolution 12 month post endoscopic laser DCR (EL-DCR) in 46 patients and the younger the age of the patient. It is well recognised that the incidence of NLDO increases with age: it is plausible to consider that an 18 score questionnaire of the GBI could be difficult to fill in for our more elderly patients, particularly, if no guidance notes are presented to them. If a simpler questionnaire can be used, which is more user friendly, a more accurate assessment of patient symptomatology can be surmised. A Finnish group have recently devised a Nasolacrimal Duct Obstruction Symptom Score (NLDO-SS), and published their results following endoscopic DCR on 64 consecutive patients.22 The questionnaire had only five items that focused on the common ocular symptoms of NLDO, and as such it was more appropriate for nasolacrimal surgery than the GBI. The authors showed a statistically significant correlation between the GBI and the NLDO-SS: although not validated, perhaps the future step would be to use this simple and sensitive tool routinely to assess symptoms post EE-DCR in an ophthalmic setting.

Mistry et al23 questioned the use of GBI in assessing the success of EE-DCR. They argued that this is a quality of life assessment and patients with anatomical obstruction at the lacrimal apparatus may not have great overall disability. Furthermore, the GBI is intended to measure change in health status, yet it is only distributed post-procedure and thus is not appropriate for assessing symptoms before surgery. The authors developed a new Lac-Q questionnaire after reviewing the presenting complaints of 100 consecutive patients referred for treatment of lacrimal obstruction. The questions account for both eye-specific symptoms as well as the overall social impact of the disease. In a pilot study of 22 DCR surgeries, they showed that compared with preoperative scores, the reduction postoperatively was significant.23 Also noted was the correlation between Lac-Q scores and objective assessments, giving merit to the use of the questionnaire in a larger study.

Our study has all the limitations associated with being retrospective in nature. The range of time from surgery to receiving questionnaire was 6–133 months and recall bias may have contributed to the results, even though the questionnaire used was validated. For future studies, the Lac-Q or NLDO-SS questionnaires can be used prospectively for patients undergoing EE-DCR.

Conclusions

EE-DCR can be considered the first-line surgical procedure for the treatment of acquired NLDO. As it improves quality of life, ophthalmologists should use measures such as GBI, Lac-Q, or NLDO-SS scores to evaluate success as an adjunct to auditing their results. Ultimately patients are not concerned with functional or anatomical outcome, but how an operation can help improve the quality of their life. Our study shows that EE-DCR gives an improvement of general well-being, as assessed by a validated questionnaire. This information should be made available to all patients before deciding their management options.