Article Text

Patients' perception of DED and its relation with time to diagnosis and quality of life: an international and multilingual survey
1. Marc Labetoulle1,
2. Maurizio Rolando2,
3. Christophe Baudouin3,
4. Gysbert van Setten4
1. 1Ophthalmology Department, Bicêtre Hospital, APHP, South Paris University, Le Kremlin-Bicêtre, France
2. 2University of Genoa, Genoa, Italy
3. 3Quinze-Vingts National Ophthalmology Hospital, Paris, France
4. 4St Eriks Eye Hospital, Stockholm, Sweden
1. Correspondence to Professor Marc Labetoulle, Ophthalmology Department, Bicêtre Hospital, APHP, South Paris University, 78 rue du Général Leclerc, Le Kremlin-Bicêtre Cedex 94275, France; marc.labetoulle{at}bct.aphp.fr

## Abstract

Background/aims To improve understanding of patients' experience and perception of dry eye disease (DED) and its impact on quality of life (QoL).

Methods This survey was observational, non-interventional and cross-sectional. The survey was conducted online on 706 patients with DED from five European countries (France, Germany, Italy, Spain and the UK). All patients met the following inclusion criteria: 40 years or older with DED diagnosed by a healthcare professional (HCP), not wearing contact lenses and using tear substitutes daily for at least 6 months. The survey (performed in the five native languages) included 9 screening questions (inclusion criteria) and 26 complementary questions about patients' demography, disease history, DED diagnosis, use of relief treatments, perceptions of DED condition and its impact on QoL.

Results Overall, 218 of 706 (31%) patients perceived DED as a ‘disease’ or even a ‘handicap’, and 468 of 706 (66%) as a ‘discomfort’. High impact of DED on patients' QoL was associated with negative perception, delay in diagnosis, visits to more than one HCP before diagnosis and high frequency of treatment use. This survey also provided us with a list of language-specific keywords that patients used most frequently to spontaneously describe their condition.

Conclusions Findings showed that negative perception of DED, delayed diagnosis and high frequency of treatment use were inter-related, and that all have a negative impact on patients' QoL. The generated language-specific keywords used to describe DED could serve as the basis for a comprehensive QoL questionnaire to be used in clinical settings.

• Ocular surface
• Treatment Medical
• Tears
View Full Text

## Introduction

Dry eye disease (DED) is a multifactorial disease affecting 6–34% of adults worldwide,1 ,2 yet it is often underdiagnosed or misdiagnosed and inadequately treated, despite the risk of ocular damage.2 ,3 The direct cost of DED has been estimated between $300 and$1100 per patient annually, with additional costs due to impaired work productivity, and economic impact on healthcare systems being similar to rheumatoid arthritis.4–6

One of the most challenging points in DED is to diagnose it early and accurately from highly variable symptoms that often overlap with other ocular surface pathologies such as allergies.7 While several frameworks attempted to standardise the process, there is still no gold standard diagnostic tool,8 ,9 particularly when considering subjective symptoms and patients' distinct ways of describing them. To the best of our knowledge, patients' words of choice when explaining discomfort or pain have not been captured in any publication yet, nor did a comparison between specific vocabularies used in the main European languages. Likewise, the assessment of DED severity is also unstandardised. Recently, the Ocular Dryness DIsease SEveritY (ODISSEY) European Consensus Group published a useful algorithm to facilitate diagnosis of severe DED based on two main criteria: corneal fluorescein staining and Ocular Surface Disease Index (OSDI).2 Nonetheless, considering symptoms alone and disregarding DED clinical signs may still lead to underdiagnosis or misdiagnosis, and therefore inefficient treatment, disease progression and lower quality of life (QoL).10

Several questionnaires (eg, OSDI,11 Symptoms Assessment iN Dry Eye (SANDE),8 Dry Eye Questionnire (DEQ),12 Canadian Dry Eye Epidemiologic Study (CANDEES),13 Ocular Surface Disease-Quality of Life (OSD-QoL)14 and Impact of Dry Eye in Everyday Life (IDEEL))15 used in clinical studies have shown DED greatly impacts patients' ability to perform everyday tasks,12 ,16 ,17 and severe DED affects QoL to a similar degree as class III/IV angina, a situation that 16% of patients would give up years of life to ameliorate.18 However, the impact of prediagnosis disease history and the influence of cultural and linguistic differences across countries on DED perception and related QoL have yet to be fully investigated.

A comprehensive online survey was conducted across five European countries (France, Germany, Italy, Spain and the UK) to capture patients' perspectives on DED in their own language and get a better understanding of their experience. The survey also examined the impact of DED on patients' QoL, their daily activities/mood and the most affecting factors.

## Materials and methods

This was an observational, non-interventional, cross-sectional, 35-question survey. All respondents gave consent to survey participation acknowledging provided information would remain anonymous and confidential, and they could discontinue it at any time.

### Survey development

The survey was designed by Listening Pharma, following preparation by three of the authors (ML, GvS and MR; online supplementary appendix A). Survey organisation and conduct was tested in face-to-face interviews with 9 volunteers selected from a panel of 6000 patients. The final English survey comprised nine screening questions to confirm inclusion criteria, followed by a 26-question survey regarding demography, comorbidities, impact on daily life, disease burden, diagnosis and treatment. The survey was translated into French, German, Italian and Spanish by Lightspeed Global Market Insight (GMI), who also distributed the survey. The questions were validated for usability and timing on a subset of respondents before being made available to the entire research panel. Post hoc analyses of the answers showed a good Cronbach's α for the entire survey (0.808) and excellent for the QoL-related questions (0.92).

### Patient population and data collection

Inclusion criteria combined a confirmed/definite DED diagnosis by a healthcare professional (HCP), daily use of artificial tears for at least 6 months, age of 40 years or older, and no contact lenses. Patients aged <40 years, who used contact lenses, whose eye condition was not diagnosed by an HCP, who had not used tear substitutes daily for at least 6 months, or who had Sjögren's syndrome, were excluded.

The nine screening questions were completed anonymously online by 52 786 volunteers selected from a pre-existing market research panel, of which 810 fulfilled the inclusion criteria. Among them, 104 patients were excluded and data analysis was performed on 706 respondents (see online supplementary table S1).

### supplementary table

The survey was available online between 25 November and 12 December 2013. It was not time-limited, and mean (SD) completion time was 20.54 (8.86) min. To avoid potential bias, responders were not aware of the survey's topic prior to participation and all information was self-reported.

### Statistical analysis

Statistical accuracy at a 95% confidence level was ±3.69% to ±8.43% based on the total sample size (N=706), the smallest country sample (Germany, N=135) and a 50% response rate. Comparisons in patient history and QoL measures between countries were examined using analysis of variance with post hoc Tukey analyses and χ² tests in Igor Pro (Wavemetrics, Lake Oswego, Oregon, USA). Frequencies responsible for significant χ² results were assessed with standardised residuals corresponding to an α of 0.05 (±1.96) according to Zar.19 Impact on seven QoL items was assessed using a scale ranging from 0=none to 10=high/extreme. Rather than comparing absolute values between countries, we decided to group patients based on a score cut-off: low impact on QoL included scores <7 and high impact scores ≥7. Time to diagnosis was calculated as the time between symptoms onset and HCP-confirmed diagnosis.

## Results

### Demographics

Differences were found between countries in patients' demography. Patients in France and the UK had a lower and higher DED duration, respectively, than the overall population (see online supplementary table S1). The presence of any chronic/acute disease was reported by 437 of 706 (62%) patients, with significant variation between countries (p<0.0001). The most commonly reported illness was ‘eye disorders or discomfort’, which was spontaneously selected as a chronic disease by 173 of 706 (25%) patients (DED diagnosis as an inclusion criterion was confirmed in a separate question). Other commonly reported diseases included chronic pain (154/706; 22%) and hypertension (151/706; 21%).

### Ocular comorbidities

Besides DED, 376 of 706 (53%) patients reported having ocular comorbidities with the most common being ‘allergies or seasonal eye inflammation’ (166/706; 24%), followed by conjunctivitis (107/706; 15%) and cataract (83/706; 12%) (see online supplementary tables S1 and S2).

### Disease history

The mean (SD) duration of DED symptoms before participation was 62.2 (70.9) months, with significant variation between countries (p=0.003) (see online supplementary table S1). Overall, 288 of 706 (41%) patients were not diagnosed at symptom onset: 197 (28%) were diagnosed within 1 year and 91 (13%) after 1 year, with no significant variation between countries.

Before diagnosis, 256 of 706 (36%) patients visited more than one HCP, with significant variation between countries (p<0.0001). Among these, 120 of 256 (47%) saw one additional HCP, 91 (36%) saw two, and 45 (18%) saw three or more. Diagnosis was performed by an ophthalmologist in 556 of 706 (79%) cases, an optician/optometrist in 84 (12%), a family doctor/general practitioner in 36 (5%) or a rheumatologist in 15 (2%).

Between symptom onset and correct diagnosis/treatment, 302 of 706 (43%) patients used one or more relief treatments: artificial tears (128/302; 43%), eye drops/ocular physiological solution (204/302; 67%) and eye cream/ointment (40/302; 14%).

### Patient descriptions of symptoms

In an open-ended question at the beginning of the survey, patients spontaneously described their eye condition using four words (table 1). In a subsequent independent question, patients chose 4 terms from a 12-option list to describe their symptoms and the top four terms were dry (443/706; 63%), itchy (301/706; 43%), foreign body sensation (294/706; 42%) and sore/irritated (273/706; 39%).

Table 1

Spontaneous terms (with their respective translations) used to describe DED*

### Negative patient's disease perception impacted QoL

The majority, 468 of 706 (66%), perceived the condition as a ‘discomfort’, 133 of 706 (19%) a ‘disease’ and 85 of 706 (12%) a ‘handicap’, with statistically significant differences between countries (p<0.0001; figure 1). The most common effects on daily life were the need to wear sunglasses (276/706; 39%), avoidance of air conditioning (217/706; 31%), disturbance in reading books for leisure (182/706; 26%), using screens for leisure (178/706; 25%) and at work (170/706; 24%). Also, 89 (13%) patients reported being woken up by their DED symptoms.

Figure 1

Patients' perception of dry eye disease (DED) varied between countries. The difference between countries in DED perception was statistically significant (p<0.0001). The symbol * shows values that were significantly lower than the overall value (p<0.05 based on standardised residual after χ2 test). Significant differences included fewer patients considering DED as a ‘disease’ in France and in the UK (14/141 or 10% and 12/145 or 8%, respectively), and fewer patients choosing the term ‘handicap’ in Italy (3/136 or 2%). The symbol † shows values that were significantly higher than the overall value (p<0.05 based on standardised residual after χ2 test). Significant differences included more patients selecting ‘handicap’ in France and Germany (30/141 or 21% and 26/135 or 19%, respectively) and more patients considering DED as a ‘disease’ in Spain (51/149 or 34%).

Patients rated the impact of DED on their daily discomfort, pain and constraints, and on personal, social, work life and global QoL. Only impact on daily constraints (p=0.001) and work life (p=0.002) were significantly different between countries.

Low/high QoL impact was related to patients' perceptions. While 210 of 706 (30%) patients reported high impact on global QoL, this percentage was significantly higher when patients considered DED a ‘disease’ or ‘handicap’, and significantly lower when they viewed DED as a ‘discomfort’ (p<0.0001; online supplementary figure S1). DED perception influenced most QoL items in a similar way, that is, more patients viewing DED as a ‘disease/handicap’ and fewer patients considering DED a ‘discomfort’ reported high impact on QoL.

### Longer time to diagnosis negatively impacted QoL

High daily discomfort (p=0.001) and pain (p=0.03) and a high impact on global QoL (p=0.005) and work life (p=0.03) were all significantly related to longer (>1 year) time to diagnosis after symptoms onset (figure 2 top).

Figure 2

Factors impacting quality of life. (Top) Longer time to diagnosis (ie, >1 year) was significantly associated with high daily discomfort (p=0.001) and daily pain (p=0.03) and with high impact on work life (p=0.03) and global quality of life (QoL) (p=0.005). (Middle) A higher frequency of daily use of relief treatments was significantly associated with more daily discomfort (p<0.0001), daily pain (p=0.0024), daily constraints (p<0.0001) and with a high impact on the work life (p=0.027), personal life (p<0.0001) and global QoL (p<0.0001). (Bottom) A longer time to diagnosis and a higher frequency of daily use of relief treatments were significantly associated with more daily discomfort (p<0.0001), daily pain (p=0.00035), daily constraints (p=0.00016) and with a high impact on the work life (p<0.0001), personal life (p=0.017) and global QoL (p<0.0001).

Time to diagnosis was also significantly associated with use of relief treatments before diagnosis (p<0.0001). Relief treatments were more commonly used by patients (64/91; 70%) who had a longer (>1 year) time to diagnosis compared with patients (238/615; 39%) with shorter time frames (≤1 year). The need to see more than one HCP before diagnosis was significantly related to impact on global QoL (p=0.0005) with a high impact for 97 of 256 (38%) patients who saw more than one HCP before diagnosis compared with 113 of 450 (25%) who saw only one HCP.

### Longer time to diagnosis and frequent use of relief treatments both negatively impacted QoL

Higher frequency in the daily use of relief treatments (≥3 times) was significantly associated with a high impact on all the QoL items, except for social life (figure 2 middle). Similarly, a longer time to diagnosis (>1 years) and a higher frequency in the daily use of relief treatments (≥3 times) were both significantly associated with a high impact on all items, except for social life (figure 2 bottom).

## Discussion

Our survey identified a relationship between delayed diagnosis, frequent use of relief treatments, negative disease perception and low QoL in patients with DED. To our knowledge, this is the first evaluation of how prediagnosis history impacted patients' lives, an aspect left out of existing questionnaires on DED-related symptoms/QoL.20 ,21 We also gathered language-specific information on keywords used by patients to spontaneously describe their ocular condition.

Screening questions ensured inclusion of only ‘definite DED cases’, that is, older patients with long-term, diagnosed, chronic DED. Clinically, they represent the most challenging population regarding diagnosis and therapeutic strategy. Conversely, as inclusion also required the daily use of artificial tears for at least 6 months and an age of 40 years or older, transient DED cases mostly occurring in younger patients (eg, following refractive surgery) were not included.22 Despite including definite cases, the majority of patients did not spontaneously describe their condition as chronic as 686 of 706 (97%) viewed it as a ‘disease/handicap/discomfort’. This suggests patients did not initially realise the ‘chronic’ status of their disease (more than 6 months according to inclusion criteria) until prompted to reflect on its impact on their lives.

Conducted on a large, varied, European representative and randomly selected patient population with DED, our survey found a longer delay (>1 year) between symptoms onset and diagnosis, and the need to see more than one HCP were closely related to low QoL. Indeed, any delay in diagnosis may lead to a worsening in symptoms and therefore, in QoL. Similarly, differences between countries' healthcare systems made it difficult to compare the impact of HCP type on disease perception. For instance, while optometrists and opticians may be unable to provide patients with DED with the right information, in Spain and the UK, they may be involved in diagnosis. In fact, our survey showed these HCPs were more frequently listed as the DED diagnosing physicians in the two countries than in others.

With DED being highly prevalent in the growing older population, early and accurate diagnosis would lead to a more effective use of all healthcare resources.4 Importantly, we found a significant relationship between delayed diagnosis (>1 year), frequent daily use of relief treatments (≥3 times) and low QoL, emphasising the need for early diagnosis and adequate therapy to improve QoL and avoid the unnecessary cost of ineffective/inappropriate medications. Early diagnosis/treatment may also prevent further disease progression and chronic inflammation that could ultimately damage the ocular surface and compromise vision.20

HCP-recorded evidence and objective assessments of clinical signs were out of the scope of this survey, since it focused on patients' self-reported data likely to reflect their views and better identify relationships between diagnosis history, disease perception and impact on QoL. As the survey results suggest, and considering the frequent discordance between DED signs and symptoms is now well accepted,23 a large-scale, retrospective chart review of both DED signs and symptoms would validate whether patients' most frequently used terms also correlated with the most accurate diagnosis. Likewise, medical records could also clarify the accuracy of patients' self-declared data. The majority of respondents (437/706; 62%) reported suffering from chronic or acute illnesses. Therefore, although patients were reminded to focus on DED only, reported information on symptoms and QoL impact could have been influenced by concomitant comorbidities. Because the survey was conducted in five European countries, the subjective responses would have been influenced by cultural and linguistic subtleties, and these differences provided helpful and relevant information to build more effective language-specific diagnostic tools. Furthermore, some intercountry variations in symptoms description and QoL impact may be explained by differences in climate and environmental factors, as discussed elsewhere.24 Similarly, different healthcare, insurance and DED management systems across countries would have also influenced disease perception (ie, ‘disease’ vs ‘handicap’, as the latter is associated with increased reimbursement in some countries). Regardless of these differences, DED was consensually found to have a major impact on QoL, and a third of patients in all European countries had a negative perception of their condition.

The unbiased way patients described their condition in our survey gives an insight for everyday clinical practice; patients should first describe their symptoms freely and HCPs, helped with a keyword list if necessary, should then obtain further details. The profile of terms used by patients to spontaneously describe their condition varied between countries, highlighting the need for country-specific and language-specific sets of descriptive words that could form the basis of a diagnostic tool for local HCPs in the future. Some terms were not DED-specific and could equally describe symptoms from other conditions (eg, allergies). The most commonly used term ‘itching’ was reported by similar proportions of patients in our survey and in epidemiological studies on seasonal allergies.25 Interestingly, certain keyword combinations (eg, ‘itching’, ‘burning’ and ‘increase of discomfort in windy conditions’) were highly evocative of DED, and a keyword-combination approach may be useful to develop future diagnostic tools.

In conclusion, this large, international and multilingual survey showed a relationship between delayed diagnosis, frequent use of relief treatments, negative disease perception and low QoL in patients with DED. The results highlight the importance of early and accurate diagnosis and treatments that can help improve daily life of patients with DED, prevent disease progression and make better use of healthcare resources.

## Acknowledgments

The authors thank Genevieve Garrigos, Santen (Evry, France), for the initiation of and continued support during this project.

View Abstract

## Footnotes

• Contributors The survey was designed by ML, MR, GvS and conducted by Véronique Zimeray, PharmD, Listening Pharma (Boulogne-Billancourt, France) who was paid for her assistance. Paid survey management and data analysis and proofreading were performed by Benjamin W Gallarda, PhD, Ogilvy CommonHealth (Paris, France). Paid editorial assistance was provided by Scinopsis Medical Writing (Fréjus, France). Santen (Evry, France) provided funding for the development of this survey and the manuscript. The authors were involved with the entire process from design to critical revision of the manuscript, and maintained complete control over the direction and content of the paper.

• Funding Santen (Evry, France) provided funding for the development of this survey and the manuscript.

• Competing interests ML is an occasional consultant for Alcon, Allergan, MSD, Santen and Théa. CB is a consultant for, or has received research grants from, Alcon, Allergan, Santen and Théa. MR is a consultant for, or has received research grants from, Alcon, Allergan, Bausch & Lomb, Santen, TRB Chemedica, Farmigea and Théa. GvS is a consultant for Santen, Théa and i.com medical GmbH, and has received honoraria as a lecturer from Santen.

• Ethics approval Société Française d'Ophtalmologie.

• Provenance and peer review Not commissioned; externally peer reviewed.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.