Introduction

Glaucoma is a progressive disease characterized by damage to the optic nerve and visual loss which, if left untreated, can lead to blindness [1,2,3]. Glaucoma may affect activities of daily living such as reading [4], walking [5], or driving [6]. Falls and inability to work are also reported [7, 8]. Approximately 10% of 70 million people with glaucoma worldwide are bilaterally blind [9].

While the aim of treatment is to prevent loss of vision and thereby preserve health-related quality of life (QoL), it remains unclear how best to measure the success of treatment. Both generic and vision-specific patient-reported outcome measures (PROMs) have not demonstrated clinically meaningful differences in QoL in recent well-designed randomized controlled trials [10, 11]. It has been proposed that existing PROMs are not sensitive enough to function as primary end points [12, 13].

Current measures of QoL tend to emphasise the effects on the patient of the disease itself [14, 15]. However, glaucoma is a largely asymptomatic disease in its early stages, so available PROMs instruments may not be able to measure what they are intended to. Therefore, it is necessary to re-evaluate what really matters to patients who are living with glaucoma.

Few studies have assessed outcome preferences among patients with glaucoma. Kulkarni et al. used focus groups to evaluate preferences among patients using intraocular pressure (IOP) lowering eye drops and patients who had undergone glaucoma surgery [16] and reported patients’ ability to maintain independent living as the most important outcome. Le et al. [17] explored perspectives among patients with newly diagnosed ocular hypertension (OHT) or mild-moderate glaucoma who might be considered for minimally invasive glaucoma surgery. Disease-related outcomes (such as the ability to drive and maintain mobility outside home) and treatment burden were expressed as important, but their relative importance was not explored. A similar but extended set of preferences was subsequently reported for surgery-naïve patients with moderate to severe glaucoma [18]. Evidence is emerging that treatment for glaucoma may generate a significant burden for patients that may negatively impact QoL, separate from the burden of disease [19].

The aim of this study is to explore outcome preferences and treatment expectations of patients across a broad range of glaucoma severity and treatment history. This information needs to be elicited from patients themselves using a robust qualitative research design. In this study, we specifically ask patients to identify those outcomes that are of utmost importance to them.

Subjects and methods

The study was designed and reported according to the Consolidated Criteria for Reporting Qualitative Research [20].

Participant recruitment and ethical approval

Patients attending glaucoma outpatient clinics at two locations in the UK, namely Moorfields Eye Hospital NHS Foundation Trust and Croydon University Hospital, were approached to participate face-to-face by the interviewer (AS). One clinic is in an urban location (central London) and the other serves a mixed urban, suburban and rural population. No relationships with participants had been established prior to study commencement. Participants had to be diagnosed with open angle glaucoma (OAG, including primary open angle glaucoma, normal-tension glaucoma and pseudoexfoliative glaucoma) or OHT and to require or have had treatment to lower IOP. Diagnosis was made by a glaucoma specialist clinician. Patients with other ophthalmic pathology such as visually significant cataract were excluded. Participants were required to be able to understand, read and speak English without translation.

Purposive sampling was used, where the participants were invited to participate based on the severity of their condition and their treatment history (Fig. 1). Specifically, we recruited participants covering a broad range of disease severity and treatment history. Ocular hypertension and early glaucoma are part of the same spectrum of disease. Both are treated and therefore it was both relevant and important to include participants with ocular hypertension in our study. Disease severity was classified for better and worse eyes separately using Hodapp-Parrish-Anderson criteria for Mean Deviation (MD) from Humphrey 24-2 visual field tests (Carl Zeiss Meditec, Dublin, CA, USA) [21, 22]. Additionally, we performed integrated visual field analyses to classify disease severity based on binocular visual function [23]. Treatment was categorized based on the maximum treatment experienced, defined by invasiveness: drops were considered to be the least invasive treatment and glaucoma surgery of any kind was considered to be the most invasive. The number of drops used was defined as the number of prescribed topical glaucoma medications.

Fig. 1: Purposive sampling covered a broad range of disease severities and treatment histories.
figure 1

Numbers of participants with corresponding disease severity and treatment history are shown. Worse eye and better eye were defined for each participant using Mean Deviation from Humphrey 24-2 visual field tests. OHT ocular hypertension.

Full size image

Sample size was not predetermined, and no minimum sample size was stipulated. Our study aims to generate data on the diversity of patient expectations. Therefore, participants were recruited and interviewed iteratively until no new concepts emerged from successive interviews (that is, data saturation was achieved) [24]. Thematic analysis was conducted on interview transcripts shortly after each interview was conducted.

In total, 25 patients were approached and agreed to participate. The refusal rate was 0%. Participant characteristics are described in Table 1.

Table 1 Characteristics of participants.
Full size table

The study was approved by the Northwest – Haydock National Health Service (NHS) Research Ethics Committee (20/NW/0347) and was conducted according to the tenets of the Declaration of Helsinki. Written informed consent was obtained from each participant prior to participation.

Interview guide

The interview guide (Fig. 2) was adapted according to our research question from that used in a recent focus group study [16]. We specifically asked participants which outcome mattered most.

Fig. 2: Topic guide.
figure 2

Used for semi-structured interviews.

Full size image

Interview procedure

Semi-structured interviews were conducted and audio-recorded between November 2020 and February 2021. Interviews were carried out after the participants’ clinical consultations, in a private room within the hospital. Some participants chose to have a family member present in the room.

Interviews were conducted face-to-face in the English language by a female PhD researcher with a background in medicine and trained in qualitative research (AS). The interview process took between 11 and 42 min and the median duration was 24 min. Prompts were used to encourage participants to expand on a point, or to clarify a question where it appeared that the participant may have misunderstood the interviewer. Field notes were made during interviews to facilitate subsequent analysis. Repeat interviews were not undertaken.

Analysis

Audio-recordings were transcribed verbatim by a professional transcription service. All transcripts were audited for accuracy, and they were not returned to participants. Transcribed data were analysed by one of the authors (AS) using thematic analysis. Themes were defined based on the study aims. The researcher grouped the data into themes and examined all the cases in the study ensuring all manifestations of each theme had been accounted for and compared. Responses to the question about the most important outcome were coded separately to eliminate bias in the analysis of responses. Nodes with more coding references than others were identified as prominent themes.

A second researcher (KH) independently read and analysed a subset of the transcripts. Discussion was conducted between the two analysts to avoid subjective judgements (reflexivity) and to solve potential limitations of different interpretation (triangulation). The second researcher also reviewed any transcripts for which the first analyst was unsure about classification. Full agreement on themes and coding was reached between the two analysts through these processes. Participant checking was not performed. The qualitative software package NVIVO 12 (QSR International, Cambridge, Massachusetts, USA) was used for data management.

Clinical information including disease severity and treatment history was extracted from patients’ medical records.

Results

We identified key domains related to OAG and OHT, relevant to patients. Data were coded and developed into 4 key themes: ‘patients’ experiences of living with glaucoma’; ‘patients’ experiences of having glaucoma treatment’; ‘most important outcomes to patients’; and ‘COVID-related concerns’. Representative direct quotations taken from interview transcripts are presented for these themes and annotated with participant codes (Table 2). Subthemes that emerged from the analysis are also presented.

Table 2 Illustrative quotations categorized by theme and subtheme.
Full size table

To identify prominent themes, we analysed the proportion of coding references associated with each theme. Patients’ experiences of having glaucoma treatment (Theme 2) and patient’s experiences of living with glaucoma (Theme 1) both received the most extensive coverage in our interviews (Fig. 3). Of note, this coverage was consistently high across patients with mild, moderate and severe glaucoma, whether defined in terms of the better-seeing eye, the worse-seeing eye or binocular visual function.

Fig. 3: Percentage of total codes per theme categorized according to disease severity in worse eye or better eye or using integrated visual field.
figure 3

A larger area indicates a greater proportion of items coded.

Full size image

Theme 1 – Patients’ experiences of living with glaucoma

Living with OAG or OHT had far-reaching effects on patients’ lives. We identified 13 distinct subthemes. Limitation of activities was at the forefront of participants’ narratives. Routine, daily-life activities, such as walking, reading, and driving were disturbed, especially for participants with severe glaucoma. Some participants were worried about the effect of glaucoma on their careers. Limitation in performing specific vision-demanding activities led to a dependent life for some patients. Another patient described how these limitations impacted their self-confidence.

Participants discussed how the nature of glaucoma and its progression affected their lives. Fear of blindness was described by some participants, including some with early glaucoma. This interacted with the lack of symptoms in the early stages of glaucoma. For example, one participant explained that there were almost no obvious symptoms, which made her question the accuracy of the diagnosis. There was variability in how disease deterioration affected patients’ lives.

Theme 2 – Patients’ experiences of having glaucoma treatment

Participants described how treatment with drops, laser, and surgery affected their daily life and productivity. Our analysis identified the most prevalent dimensions of treatment burden among participants.

A commonly reported consideration among patients with a history of surgical treatment was the postoperative period. One participant explained that she was worried that her operated eye would be affected because of the heat from cooking. Laser, on the other hand, was quick and convenient for some patients, with some reporting transient discomfort after the procedure. Some patients who had taken drops were concerned about drop-related adverse events and wanted to avoid them. They raised the issues of systemic effects such as overdosing and allergic reactions.

The perceived burden of treatment also included the task of understanding treatments and appraising them based on their intensity and stability. These factors interacted with each other. For example, patients often described having to learn names of drops, understand the medications used, and determine how long the treatment would last with sometimes limited information from clinicians. Lastly, attending multiple hospital appointments was also a problem in glaucoma care.

Theme 3 – Most important outcomes to patients

Our analysis identified 6 distinct subthemes for the most important outcomes among participants. All quotations and themes emerged from asking patients directly about the most important outcomes following the treatments they had received throughout their care.

We categorized 3 of these subthemes (IOP control, maintaining vision, and being independent) as outcomes related to disease and 3 of these subthemes (treatment that does not change, freedom from eye drops, and one-time treatment) as outcomes related to treatment.

Most important outcomes related to disease

IOP control

Many participants cited IOP control, though some participants did not associate it with noticeable vision changes. Contrary to the majority, some participants expressed the view that IOP control was not important (Table 2). Others reported that they did not understand how IOP may be linked to blindness.

Maintaining vision

Patients in our study cohort across a range of ages and disease stages expressed that maintaining current vision was of utmost importance. Specifically, participants whose vision had been severely affected by OAG were worried about deteriorating vision and were willing to consider any treatment to avoid losing their vision.

Being independent

The burden of visual loss among participants with severe OAG extended to inability to commute and other hallmarks of independence. Some participants with severe OAG relied on others, which made them feel dependent. Interestingly, participants with either early or severe OAG were concerned about the efforts they made to comply with medications. For instance, a participant with mild OAG explained that his wife always assisted him when instilling eye drops.

Most important outcomes related to treatment

Treatment that does not change

Participants often had been using glaucoma drops for a long time with multiple changes.

Freedom from eye drops

Participants liked convenient and simple treatments, with freedom from drops being considered preferable. The amount of time and cognitive effort needed to maintain their treatment was also noted.

One-time treatment

Some participants did not consider ongoing treatment with eye drops as a success. One patient hoped that a new treatment would be discovered as a one-time treatment to stabilize her disease.

Theme 4 – COVID-related concerns

This study was conducted during the COVID-19 pandemic in the UK. The limited mobility outside of the home led to feelings of worry, worsened by concerns about access to eye care services. Cancelled hospital appointments and limited access to pharmacy to get eye drops were a source of stress and anxiety.

Discussion

This study explored outcomes that matter to patients diagnosed with glaucoma or OHT. To our knowledge, this is the first study to report patients’ preferences across a wide range of glaucoma severities and treatments. Our findings highlight how outcomes related both to disease and to treatment are important to patients. We planned from the outset to ask patients directly what they considered to be the most important outcomes. Using this method, we identified 6 outcomes of prime importance, of which 3 were disease-related and 3 were treatment-related.

This work builds on recent evidence about the burden of treatment in glaucoma and its negative impact on QoL [19]. We now demonstrate patients’ unequivocal interest in treatment-related outcomes, supporting the hypothesis that QoL in glaucoma may be influenced by the burden of its treatment.

The European Glaucoma Society Guidelines state that the goal of care for people with glaucoma is to promote their well-being and quality of life [25]. It is important to distinguish patients’ expectations from those of the clinicians who look after them because they may differ [26]. We were particularly careful to avoid bias from overlaying analysts’ personal opinions during scrutiny of transcripts by asking patients specifically what they considered to be the most important outcome. Only these items were included in Theme 3 (most important outcomes). For example, clinicians may prioritize slowing down disease progression [12] and patients did indeed show appreciation of disease stability. However, no patient actually identified it as the most important aim of glaucoma treatment from their perspective.

Interestingly, our study has shown that IOP is considered as the most important outcome by some patients. This is in keeping with previous work which ranked patient’s preferences [27, 28]. However, our work suggests that patients may also have different expectations to each other. The reasons for this need to be explored.

There is no clear consensus on whether the better-seeing eye or the worse-seeing eye has greater influence on vision-related quality of life [21]. Some investigators suggest that the MD of the better-seeing eye overestimates the impact of visual field loss and argue in favour of assessing binocular visual function [23, 29]. Others have reported the exact opposite [30]. Importantly, we found extensive coverage of both disease-related and treatment-related issues across the full range of glaucoma severity regardless of whether this was defined in terms of the better-seeing eye, worse-seeing eye, or binocular visual function. Therefore, our conclusion is unaffected by the ongoing debate about how to assess glaucoma severity.

This study was conducted in the midst of the COVID-19 pandemic. We observed that COVID-related concerns were mainly about limited access to eye care. Several participants expressed a fear of going to clinic. Yet, no participant identified these concerns as being of prime importance on direct questioning. We therefore think it is unlikely that patient expectations shifted due to the pandemic. Regardless, the presently identified themes are more likely to be of ongoing relevance than themes identified by investigators prior to the pandemic.

One of the strengths of this study is that it used purposive sampling of patients combined with a recruitment endpoint determined by saturation. These techniques together enabled us to generate rich responses from participants with diverse demographics, disease statuses and treatment histories [31]. For example, participants with OHT with a previous history of minimally invasive glaucoma surgery were recruited in our study. While this is not the commonest scenario for patients with OHT, we deliberately included such cases to ensure that everyone’s experiences and preferences were represented. Thus, our findings are comprehensive, not biased. By contrast, previous studies chose cohorts of patients with narrowly-defined disease severities and treatment histories [17, 18]. For example, Bicket et al. excluded the views of patients who had already experienced surgery [18]. In another study, the majority of recruits were Caucasians above 60 years of age [16], so it may not have captured adequately the views of younger patients and those of other races. The heterogeneity observed in our cohort allowed us to generate information-rich data and explore outcomes from a broad range of perspectives. For example, we found that some patients with early glaucoma were affected by fear of going blind, lending support to the idea that the mere diagnosis of a potentially blinding disease may impact quality of life in the absence of actual major vision loss [32].

We chose one-on-one interviews rather than focus groups for our study. Articulate, confident and motivated individuals contribute effectively to focus group discussions, but such studies may inadvertently exclude valid views of less articulate participants.

In terms of comorbidity, we ensured that participants’ views were not contaminated as result of non-glaucomatous ocular comorbidities. In one recent study, forty percent of patients had cataract in both eyes [17]. Since cataract is a cause of vision impairment distinct from glaucoma, identified themes of vision-dependent activities of daily living and problems with general visual function may have been contaminated by the influence of cataract. In our study, patients with non-glaucomatous causes of vision impairment were excluded to ensure that elicited preferences were related only to glaucoma.

Limitations

We recognise some limitations to our study. We recruited from only two hospital centres in the UK, which may limit the extent to which our findings are generalisable. Nonetheless, we have ensured that a broad range of perspectives were covered by recruiting a diverse sample consisting of participants of various ages, genders, ethnicities, disease profiles and treatment histories, as demonstrated in Table 1. Moreover, by recruiting from two locations serving urban, suburban and rural populations, we aimed to include the full spectrum of social diversity across the UK. In common with previous qualitative studies, this permits a degree of conceptual transference in UK glaucoma patients under NHS care [33, 34].

Our findings are not statistically representative of the entire population. However, the aim of this study is to capture the range of patient preferences by maximizing variation in the sample. We ensured that the full diversity of attitudes was represented by continuing to recruit participants until thematic saturation was established. It is well recognized that thematic saturation can be reached with relatively modest numbers of participants [24]. In this study, saturation was reached following interviews with 25 patients. Whereas qualitative studies are ideal for in-depth exploration of participants’ attitudes, we acknowledge that it is not possible to draw quantitative inferences from this type of work. Whether preferences are stable with time for individual patients warrants further investigation.

Conclusion

This study brings new perspectives on outcomes that are valued by patients who have glaucoma by directly exploring their experiences, future expectations and priorities. Glaucoma places unique burdens on patients, and we show that these can be related not only to the disease itself but also to its treatment. This needs to be considered in the further development of glaucoma PROMs if they are to measure quality of life successfully.

Summary

What was known before

  • Preserving health-related quality of life (QoL) is the ultimate therapeutic goal in glaucoma management.

  • Existing patient-reported outcome measures (PROMs) used to evaluate QoL in recent clinical trials have been criticised for being insufficiently sensitive to capture changes in health status.

  • Current measures of QoL tend to emphasise the effects on the patient of the disease itself.

What this study adds

  • By directly exploring patients’ outcome preferences, we found that patients care not only about the disease itself but also about the burden of its treatment.

  • Both treatment-related and disease-related outcomes may need to be assessed when evaluating quality of life in glaucoma.