Advertisement

Charles Bonnet syndrome: development and validation of a screening and multidimensional descriptive questionnaire

Open AccessPublished:September 03, 2018DOI:https://doi.org/10.1016/j.jcjo.2018.05.008

      ABSTRACT

      Objective

      The purpose of this study was to develop a French Canadian questionnaire for the detection of Charles Bonnet syndrome that allows for (i) valid screening and (ii) the examination of different dimensions of the client's visual hallucinations in order to better assess the resulting needs.

      Method

      Questionnaire development was guided by interviews with visually impaired individuals experiencing visual hallucinations, as well as supported by scientific literature and expert experience. A clinical study involving 76 individuals with low vision was conducted to determine the sensitivity and specificity of the instrument according to criterion validation.

      Results

      Of the 54 closed-ended questions, a subset of 11 revealed a sensitivity of 1.00 and a specificity of 0.77. Two additional questions showed high discriminating potential. Improvements to the wording and structure of some questions aiming at needs assessment were identified and applied. The improved version consists of 55 questions grouped in 8 dimensions: (1) Screening; (2) Characteristics of hallucinations; (3) Psychological impact; (4) Psychopathological origin; (5) Coping strategies; (6) Context of appearance of hallucinations; (7) Time-related matters; (8) Psychosocial support. The screening is operationalized through an algorithm applied to the set of 13 questions.

      Conclusion

      The questionnaire will be a valuable aid in screening for Charles Bonnet syndrome among the low vision clientele. However, the screening will need to be supplemented by a focused low vision interdisciplinary assessment including a visual examination and a clinical interview with a psychologist.

      Résumé

      Objectif

      La présente étude avait pour objectif la création d'un questionnaire permettant le repérage du syndrome de Charles Bonnet à l'intention des Canadiens d'expression française. Nous souhaitions (i) réaliser un repérage valide et (ii) examiner les différentes dimensions des hallucinations visuelles pour mieux évaluer les besoins des sujets atteints.

      Méthodes

      Des entrevues réalisées auprès de personnes qui présentaient à la fois une perte de vision et des hallucinations visuelles ont permis l’élaboration du questionnaire, qui a également été étayé par la littérature scientifique et des experts. Une étude clinique qui regroupait 76 personnes présentant un déficit visuel a été menée afin de déterminer la sensibilité et la spécificité de l'outil en fonction des critères de validation.

      Résultats

      Onze des 54 questions fermées avaient une sensibilité de 1,00 et une spécificité de 0,77. Deux autres questions présentaient un degré élevé de potentiel de discrimination. Des améliorations ont été apportées à la formulation et à la structure de certaines questions sur l’évaluation des besoins. La version améliorée compte 55 questions regroupées en 8 dimensions: (1) repérage; (2) caractéristiques des hallucinations; (3) impact psychologique; (4) origine psychopathologique; (5) stratégies d'adaptation; (6) contexte d'apparition des hallucinations; (7) temporalité; et (8) soutien psychosocial. Le repérage est opérationnalisé grâce à un algorithme appliqué aux 13 questions de la dimension 1 (repérage).

      Conclusion

      Le questionnaire sera d'une grande utilité pour repérer le syndrome de Charles Bonnet chez des sujets en basse vision. Cela dit, il convient d'associer à ce repérage un examen interdisciplinaire approfondi axé sur la perte de vision qui comprendra un examen de la vue et une entrevue clinique avec un psychologue.
      Apart from hallucinations associated with psychiatric, pharmacological, toxicological, or neurological problems, visual hallucinations (VH) may occur as a result of partial or complete vision loss.
      • Holroyd S
      • Rabins PV
      • Finkelstein D
      • Nicholson MC
      • Chase GA
      • Wisniewski SC
      Visual hallucinations in patients with macular degeneration.
      • Menon GJ.
      Complex visual hallucinations in the visually impaired: a structured history-taking approach.
      • Menon GJ
      • Rahman I
      • Menon SJ
      • Dutton GN
      Complex visual hallucinations in the visually impaired: the Charles Bonnet syndrome.
      • Schadlu AP
      • Schadlu R
      • Shepherd 3rd, JB
      Charles Bonnet syndrome: a review.
      The composition of the VH varies among people, but they may be categorized as simple (e.g., geometric shapes) or complex (faces, people, objects, etc.).
      • Aziz VM
      • Nessim M
      • Warner NJ
      Charles-Bonnet syndrome and musical hallucination.
      • Tien AY
      Distributions of hallucinations in the population.
      A clinical presentation known as Charles Bonnet syndrome (CBS) has been further studied. It usually refers to complex visual hallucinations that occur in the presence of vision loss and in the absence of psychiatric disorders.
      • Schadlu AP
      • Schadlu R
      • Shepherd 3rd, JB
      Charles Bonnet syndrome: a review.
      • Jackson ML
      • Bassett K
      • Nirmalan PK
      Charles Bonnet hallucinations: natural history and risk factors.
      A key characteristic is that individuals experiencing CBS are fully or partially aware that they are hallucinating.
      • Menon GJ.
      Complex visual hallucinations in the visually impaired: a structured history-taking approach.
      ,
      • Ffytche DH.
      Visual hallucinatory syndromes: past, present, and future.
      • Pang L.
      Visual hallucinations: identifying Charles Bonnet syndrome.
      • Russell G
      • Harper R
      • Allen H
      • Baldwin R
      • Burns A
      Cognitive impairment and Charles Bonnet syndrome: a prospective study.
      Some cases of CBS may present with only simple hallucinations.
      • Khan JC
      • Shahid H
      • Thurlby DA
      • Yates JRW
      • Moore AT
      Charles Bonnet syndrome in age-related macular degeneration: the nature and frequency of images in subjects with end-stage disease.
      • Nalcaci S
      • Ilim O
      • Oztas Z
      • et al.
      The prevalence and characteristics of Charles Bonnet syndrome in Turkish patients with retinal disease.
      Studies suggest that the proportion of people with visual impairment who experience VH ranges from 5% to 27%.
      • Menon GJ.
      Complex visual hallucinations in the visually impaired: a structured history-taking approach.
      • Schadlu AP
      • Schadlu R
      • Shepherd 3rd, JB
      Charles Bonnet syndrome: a review.
      ,
      • Abbott EJ
      • Connor GB
      • Artes PH
      • Abadi RV
      Visual loss and visual hallucinations in patients with age-related macular degeneration (Charles Bonnet syndrome).
      • Lannon SP
      • Stevenson MR
      • White ST
      • Logan JF
      • Reinhardt-Rutland AH
      • Jackson AJ
      Visual hallucinations in patients with age-related macular degeneration (AMD).
      The stigmatization and fear of confiding or of being considered mentally ill by relatives or health professionals could hide a much higher prevalence of VH.
      • Menon GJ.
      Complex visual hallucinations in the visually impaired: a structured history-taking approach.
      • Menon GJ
      • Rahman I
      • Menon SJ
      • Dutton GN
      Complex visual hallucinations in the visually impaired: the Charles Bonnet syndrome.
      • Schadlu AP
      • Schadlu R
      • Shepherd 3rd, JB
      Charles Bonnet syndrome: a review.
      ,
      • Nair AG
      • Nair AG
      • Shah BR
      • Gandhi RA
      Seeing the unseen: Charles Bonnet syndrome revisited.
      With the aging of the population, more people will experience vision loss, and the prevalence of CBS will increase accordingly.
      • Pang L.
      Hallucinations experienced by visually impaired: Charles Bonnet syndrome.
      Thus, some of the people who will experience visual impairment (VI) will also have to deal with VH due to CBS. For someone living with VH, learning that their hallucinations are not the result of mental illness can be reassuring.
      • Menon GJ
      • Rahman I
      • Menon SJ
      • Dutton GN
      Complex visual hallucinations in the visually impaired: the Charles Bonnet syndrome.
      • Lannon SP
      • Stevenson MR
      • White ST
      • Logan JF
      • Reinhardt-Rutland AH
      • Jackson AJ
      Visual hallucinations in patients with age-related macular degeneration (AMD).
      ,
      • Nadarajah J.
      Visual hallucinations and macular degeneration: an example of the Charles Bonnet syndrome.
      • Gordon KD
      • Felfeli T
      Family physician awareness of Charles Bonnet syndrome.
      In addition, a correct diagnosis leads to a better assessment of needs and more appropriate patient management.
      • Nair AG
      • Nair AG
      • Shah BR
      • Gandhi RA
      Seeing the unseen: Charles Bonnet syndrome revisited.
      • Gordon KD
      • Felfeli T
      Family physician awareness of Charles Bonnet syndrome.
      ,
      • O'Farrell L
      • Lewis S
      • McKenzie A
      • Jones L
      Charles Bonnet syndrome: a review of the literature.
      However, a survey performed in Canada with family physicians found that about 75% of respondents were not at all or slightly aware of Charles Bonnet syndrome and that 85% of respondents never discussed the possibility of CBS with their patients who present low vision.
      • Gordon KD
      • Felfeli T
      Family physician awareness of Charles Bonnet syndrome.
      It is therefore essential to facilitate the identification of CBS, particularly in the field of rehabilitation for visual impairment.
      • O'Farrell L
      • Lewis S
      • McKenzie A
      • Jones L
      Charles Bonnet syndrome: a review of the literature.
      However, there is currently no validated tool available to screen for CBS. This led our team to develop and validate a French Canadian questionnaire. This questionnaire is not only designed to screen for CBS, but it also allows for the examination of different dimensions of the client's visual hallucinatory manifestations in order to better assess the intervention needs.

      Method

      The research protocol was approved by the Research Ethics Board of the Centre for Interdisciplinary Research in Rehabilitation of Greater Montreal (CRIR-868-0713). Informed consent to participate was obtained from all subjects.

      Clinical expertise

      The expertise of a group of clinicians in low vision rehabilitation at the Institut Nazareth et Louis-Braille du CISSS de la Montérégie-Centre was brought together. With an interdisciplinary perspective, 5 clinicians were involved: 2 psychologists, an optometrist, a social worker, and a specialist in clinical activities, all familiar with persons experiencing VH.

      Procedure

      Figure 1 summarizes the 3 steps that were followed to develop the Questionnaire de repérage du syndrome de Charles Bonnet (QR-SCB; in English, Charles Bonnet Syndrome Screening Questionnaire) and highlights the main achievements associated with each step.
      Fig 1
      Fig. 1Steps for the development of the questionnaire.
      To identify the topics to be included in the questionnaire (step 1), an interview guide was developed. Based on a literature review and the consultation of experts, it included 10 open-ended questions. Interviews were conducted with 10 visually impaired persons aged between 69 and 81 years who were experiencing VH. The content analysis of the interviews led to step 2: the development of the questionnaire structure and the formulation of questions. A pretest with 41 participants with VI, aged between 55 and 90 years, gave rise to the first version of the questionnaire, which comprised 54 closed-ended questions. Step 3 focused on the criterion validation executed on this first version.

      Criterion validation

      A clinical study was conducted to estimate the sensitivity and specificity of the QR-SCB. A criterion validation methodology was used, where the criteria consisted in the clinicians’ judgments of the presence/absence of CBS in participants. Seventy-eight participants with a visual acuity of less than 0.2 logMAR (equivalent to 6/9.5 [20/32] on an Early Treatment Diabetic Retinopathy Study chart) were recruited at the Institut Nazareth et Louis-Braille du CISSS de la Montérégie-Centre by the Front Line-Evaluation-Orientation staff or by other clinicians, or externally by recruitment announcements. The sample consisted of 48 women and 30 men aged between 48 and 102 years (mean 72.3 years, SD 11.9 years). The most common diagnosis was age-related macular degeneration. The Montreal Cognitive Assessment, adapted for people with VI (MoCA-Blind),
      • Wittich W
      • Phillips N
      • Nasreddine ZS
      • Chertkow H
      Sensitivity and specificity of the Montreal Cognitive Assessment modified for individuals who are visually impaired.
      which aims to detect individuals at risk for mild cognitive impairment, was administered to ensure the absence of any dysfunction of this type. This measure has previously been successfully administered in the context of studying individuals with and without CBS.
      • Boxerman H
      • Wittich W
      • Overbury O
      Charles Bonnet syndrome in older adults with age-related macular degeneration: its relationship to depression and mild cognitive impairment.
      Four research assistants, 2 optometrists, and 2 psychologists participated in the data collection. In a single 2−3 hour session, each participant first met with a trained research assistant for the verbal administration of the QR-SCB. Then, each participant met with an optometrist or a psychologist, or both. The optometrist performed an ocular examination of the participant and the psychologist conducted a focused clinical interview to identify unusual visual perceptions. This interview was based on a semistructured interview following Menon's guide.
      • Menon GJ
      • Rahman I
      • Menon SJ
      • Dutton GN
      Complex visual hallucinations in the visually impaired: the Charles Bonnet syndrome.
      When the participant met only 1 clinician, the clinician judged the presence or absence of CBS alone. When both clinicians met with the participant, they met at the end of the session to share their respective clinical observations in order to reach a consensus judgment on the case. Clinicians made their judgment without being aware of the answers collected with the QR-SCB by the research assistant.

      Data analysis

      Of the 54 questions in the questionnaire, 22 were thought to have some potential to discriminate cases of CBS (as opposed to questions for assessing the client's experience). The analysis of the discriminating power of these 22 questions was carried out in 3 stages:
      • (1)
        A careful comparison, question by question, between the answers obtained from the participants identified by the clinicians as presenting with CBS and those without this syndrome.
      • (2)
        χ2 tests on the questions retained after the previous examinations.
      • (3)
        The search, by trial and error, for the optimal combination of questions in terms of sensitivity and specificity, among the questions retained after the previous operations.

      Results

      Quantitative

      Figure 2 displays the breakdown of participants according to the questionnaire screening and the external criterion as expressed by the clinical judgment.
      Fig 2
      Fig. 2Breakdown of participants according to the screening by the QR-SCB and the external criterion.
      Two of the 78 cases were excluded from the analyses: one because of the ambiguity of the case (no consensus judgment obtained) and the other because of the participant's difficulties to understand the questions. The distribution of the 76 remaining cases is as follows: optometrist, n = 21; psychologist, n = 24; optometrist and psychologist jointly, n = 31. All judgments combined, clinicians identified 21 cases of Charles Bonnet syndrome.
      Sensitivity, specificity, and discriminating power. Nine of the 22 potentially discriminating questions provided no power to discriminate cases and were therefore eliminated. Among the remaining 13 questions, the best indices of sensitivity and specificity were obtained with a group of 11 questions simply aimed at identifying the nature of unusual visual perceptions by distinguishing simple hallucinations (e.g., lines, circles, patterns) and complex ones (e.g., people, landscapes, trees) from an entoptic phenomenon (e.g., pulsations or light flickers) on a 4-level frequency scale (never to always).
      The sensitivity of the 11-item subset is 1.00. In other words, it identified 100% of those with CBS (n = 21/21). A specificity index of 0.77 was obtained for this group of questions when the external criterion consisted of a joint judgment by the optometrist and the psychologist. Thus, this subset of questions correctly identified 77% of people who did not present CBS (20/26). The indices of specificity obtained according to the other modalities of the clinical judgment (external criterion) are 0.42 for the optometrist alone (5/12) and 0.47 (8/17) for the psychologist alone. Table 1 presents the contingency tables for the three modalities of the external criterion.
      Table 1Contingency tables for the 3 modalities of the external criterion
      Clinical Judgments
      Optometrist + Psychologist

      (sensitivity = 0.77; n = 31)
      Psychologist

      (sensitivity = 0.47; n = 24)
      Optometrist

      (sensitivity = 0.42; n = 21)
      With CBSWithout CBSWith CBSWithout CBSWith CBSWithout CBS
      QR-SCB
       Positive567997
       Negative0200805
      CBS, Charles Bonnet syndrome; QR-SCB, Charles Bonnet Syndrome Screening Questionnaire.
      The remaining 2 questions asked the participant, on the same 4-level frequency scale, if his visual hallucinations were clearer than what his vision currently allows him and if they were clear enough to look real. The χ2 tests show significant differences between the true positive cases (with CBS, n = 21) and the false positive cases (entoptic phenomena, n = 13) who answered both questions (question no. 12: χ2 [1, n = 34] = 6.05, p = 0.01; and question no. 13: χ2 [1, n = 34] = 4.114, p = 0.04).

      Qualitative

      The administration of the QR-SCB to the participants in the clinical study took between 15 and 45 minutes. It revealed that qualitative improvements, like formulation and choice of answers, were required with some of the questions aimed at assessing the client's experience with his visual hallucinations (that is, those not used for screening) as well as to certain sequences of questions. However, the object of the questions remained the same between the two versions of the questionnaire.

      Discussion

      Aside from Menon's interview guide,
      • Menon GJ.
      Complex visual hallucinations in the visually impaired: a structured history-taking approach.
      there is currently no validated tool to identify CBS and to describe various dimensions of the person's experience with it. Therefore, the purpose of the present study was to fill this gap. Menon's guide comprises 13 open-ended questions allowing for the detection of unusual visual perceptions. In comparison, the QR-SCB presented here uses closed-ended questions. In this study, 11 of these questions identified all cases of CBS. Based on the answers to these questions on a 4-level frequency scale, the screening is operationalized through an algorithm. This makes the questionnaire easier to administer and interpret, even in the absence of clinical experience or knowledge of CBS.
      With a sensitivity of 1.00 (for all modalities of the external criterion) and a specificity of 0.77 (when the external criterion is a clinical joint judgment), the QR-SCB shows respectable metrological qualities. The specificity index would probably have been higher if the 2 other questions showing discriminatory power had been included in the calculation. Since not every participant answered these 2 questions, such a calculation was not possible. However, a simulation showed that among the 13 false positive participants who answered both questions, 9 would have been identified as true negatives (without VH) if these 2 questions had accompanied the 11-item set previously identified as discriminant. While waiting for a more thorough demonstration of the real added value of these 2 questions to the specificity of the QR-SCB, our results suggest that, in any use of the questionnaire, they should be considered in the same way as the 11 discriminating questions. It must also be mentioned that there were not enough cases of VH from other etiologies to verify the questionnaire's power to discriminate adequately between them and CBS cases.
      Since CBS cases are most commonly seen in the elderly with acquired visual impairment,
      • Menon GJ
      • Rahman I
      • Menon SJ
      • Dutton GN
      Complex visual hallucinations in the visually impaired: the Charles Bonnet syndrome.
      it is important to make sure that the questionnaire shows comprehensibility and fluidity, not only to facilitate the identification of CBS, but also to properly identify the client's intervention needs. This second version comprises 55 questions, 1 more than the previous version (1 question was split in 2), that are organized in 8 dimensions:
      • (1)
        Screening (11 questions)
      • (2)
        Characteristics of hallucinations (6 questions)
      • (3)
        Psychological impact (4 questions)
      • (4)
        Psychopathological origin (6 questions)
      • (5)
        Coping strategies (9 questions)
      • (6)
        Context of appearance of hallucinations (9 questions)
      • (7)
        Time-related matters (5 questions)
      • (8)
        Psychosocial support (5 questions)
      The improvements brought into the second version will probably make it easier to administer, as well as reduce its administration time. To date, feedback from the clinicians who administered the revised QR-SCB indicates the usefulness of the section focusing on the assessment of the client's experience with his visual hallucinations (dimensions 2 to 8). They report that when required, the picture provided by this part of the questionnaire allows for the development of more targeted intervention plans, better adapted to the needs of the individual. The questionnaire is available online at: http://www.inlb.qc.ca/professionnels-recherche/recherche/qr-scb/.

      Conclusion

      In its current version, the QR-SCB presented here appears to be an interesting and clinically useful instrument for the identification of visual hallucinations in persons with visual impairment in the absence of mental disorders as well as for planning interventions. However, the screening performed by the discriminating subset of questions has to be complemented by a targeted low vision interdisciplinary assessment that crosses observations obtained from a visual examination and those from a clinical interview with a psychologist. Additional research efforts are desirable 1) to better understand the metrological qualities of this tool, including its 11 ability to discriminate VH from other etiologies and 2) to assess the effect of the use of the QR-SCB on the clients' care path and/or intervention plans.

      Footnotes and Disclosure

      The authors have no proprietary or commercial interest in any materials discussed in this article.

      Supported by

      This work was funded by the fund of the Institut Nazareth et Louis-Braille (INLB) du CISSS de la Montérégie-Centre and by the Vision Health Research Network (VHRN) of the Quebec Health Research Fund. The INLB conducted the whole research while the VHRN had no involvement in its conduct.

      Acknowledgements

      The authors warmly thank Francine Baril, documentation technician at INLB; the participants at the different stages of the research, as well as Roger Dufour, Louis Macchabée, Catherine Bergeron, Cynthia Dutrisac, Jean-Philippe Quenneville, Wanseo Kim, and Fanie Chainey, who participated in data collection and/or analysis.

      References

        • Holroyd S
        • Rabins PV
        • Finkelstein D
        • Nicholson MC
        • Chase GA
        • Wisniewski SC
        Visual hallucinations in patients with macular degeneration.
        Am J Psychiatry. 1992; 149: 1701-1706
        • Menon GJ.
        Complex visual hallucinations in the visually impaired: a structured history-taking approach.
        Arch Ophthalmol. 2005; 123: 349-355
        • Menon GJ
        • Rahman I
        • Menon SJ
        • Dutton GN
        Complex visual hallucinations in the visually impaired: the Charles Bonnet syndrome.
        Surv Ophthalmol. 2003; 48: 58-72
        • Schadlu AP
        • Schadlu R
        • Shepherd 3rd, JB
        Charles Bonnet syndrome: a review.
        Curr Opin Ophthalmol. 2009; 20: 219-222
        • Aziz VM
        • Nessim M
        • Warner NJ
        Charles-Bonnet syndrome and musical hallucination.
        Int Psychogeriatr. 2004; 16: 489-491
        • Tien AY
        Distributions of hallucinations in the population.
        Soc Psychiatry Psychiatr Epidemiol. 1991; 26: 287-292
        • Jackson ML
        • Bassett K
        • Nirmalan PK
        Charles Bonnet hallucinations: natural history and risk factors.
        Int Congr Ser. 2005; 1282: 592-595
        • Ffytche DH.
        Visual hallucinatory syndromes: past, present, and future.
        Dialogues Clin Neurosci. 2007; 9: 173-189
        • Pang L.
        Visual hallucinations: identifying Charles Bonnet syndrome.
        Int J Ophthalmol Eye Sci. 2015; S1: 14-22
        • Russell G
        • Harper R
        • Allen H
        • Baldwin R
        • Burns A
        Cognitive impairment and Charles Bonnet syndrome: a prospective study.
        Int J Geriatr Psychiatry. 2018; 33: 39-46
        • Khan JC
        • Shahid H
        • Thurlby DA
        • Yates JRW
        • Moore AT
        Charles Bonnet syndrome in age-related macular degeneration: the nature and frequency of images in subjects with end-stage disease.
        Ophthalmic Epidemiol. 2008; 15: 202-208
        • Nalcaci S
        • Ilim O
        • Oztas Z
        • et al.
        The prevalence and characteristics of Charles Bonnet syndrome in Turkish patients with retinal disease.
        Ophthalmologica. 2016; 236: 48-52
        • Abbott EJ
        • Connor GB
        • Artes PH
        • Abadi RV
        Visual loss and visual hallucinations in patients with age-related macular degeneration (Charles Bonnet syndrome).
        Invest Ophthalmol Vis Sci. 2007; 48: 1416-1423
        • Lannon SP
        • Stevenson MR
        • White ST
        • Logan JF
        • Reinhardt-Rutland AH
        • Jackson AJ
        Visual hallucinations in patients with age-related macular degeneration (AMD).
        Vis Impair Res. 2006; 8: 9-16
        • Nair AG
        • Nair AG
        • Shah BR
        • Gandhi RA
        Seeing the unseen: Charles Bonnet syndrome revisited.
        Psychogeriatrics. 2015; 15: 204-208
        • Pang L.
        Hallucinations experienced by visually impaired: Charles Bonnet syndrome.
        Optom Vis Sci. 2016; 93: 1466-1478
        • Nadarajah J.
        Visual hallucinations and macular degeneration: an example of the Charles Bonnet syndrome.
        Aust N Z J Ophthalmol. 1998; 26: 63-65
        • Gordon KD
        • Felfeli T
        Family physician awareness of Charles Bonnet syndrome.
        Fam Pract. 2018 Feb 17; ([epub ahead of print])
        • O'Farrell L
        • Lewis S
        • McKenzie A
        • Jones L
        Charles Bonnet syndrome: a review of the literature.
        J Vis Impair Blind. 2010; 104: 261-274
        • Wittich W
        • Phillips N
        • Nasreddine ZS
        • Chertkow H
        Sensitivity and specificity of the Montreal Cognitive Assessment modified for individuals who are visually impaired.
        J Vis Impair Blind. 2010; 104: 360-368
        • Boxerman H
        • Wittich W
        • Overbury O
        Charles Bonnet syndrome in older adults with age-related macular degeneration: its relationship to depression and mild cognitive impairment.
        Br J Vis Impair. 2015; 33: 19-30