Quality of life of eye amputated patients

Authors


Marie Louise Roed Rasmussen, MD
University of Copenhagen
Department of Neuroscience and Pharmacology
Eye Pathology Institute
Frederik den V’s vej 11
DK 2100 Copenhagen
Denmark.
Tel: + 45 35 32 60 83
Fax: + 45 35 45 60 80
Email: dr.roed@gmail.com

Abstract.

Purpose:  To evaluate eye-amputated patients’ health-related quality of life, perceived stress, self-rated health, job separation because of illness or disability and socioeconomic position.

Methods:  Patients were recruited from a tertiary referral centre situated in Copenhagen. Inclusion criteria were eye amputation, i.e. evisceration, enucleation, orbital exenteration or secondary implantation of an orbital implant during the period 1996–2003, and participation in a previous investigation (2005).

In total, 159 eye-amputated patients were included, and completed a self-administered questionnaire containing health-related quality of life (SF-36), the perceived stress scale and answered questions about self-rated health, job changes because of illness or disability and socioeconomic status. These results were compared with findings from the Danish Health Interview Survey 2005.

Results:  The eye-amputated patients had significantly (p < 0.05) lower scores (poorer health) on all SF-36 subscales and more perceived stress compared to the general population. In all, 43.3% of the patients rated their health as excellent or very good compared to 52.1% of the general population. In total, 25% of the study population has retired or changed to a part-time job because of eye disease.

The percentage of eye amputated patients, who were divorced or separated, was twice as high as in the general population.

Conclusion:  The impact of an eye amputation is considerable. The quality of life, perceived stress and self-rated health of many eye-amputated patients are drastically changed. Eye amputation has a marked negative influence on job separation because of illness or disability and on socioeconomic position.

Introduction

Eye amputation (EA) is often the endpoint of a long clinical journey. Eye amputation patients vary in age, have different consecutive diagnoses and indications for eye amputation and different surgical treatments (Rasmussen et al. 2010). After their final surgeries, however, they share visual impairment and facial disfigurement. This brings about dramatic changes in many aspects of the EA patients’ lives.

The health-related quality of life (HRQOL) of anophthalmic patients has recently been reported for a group of Korean patients. These patients had significantly lower HRQOL scores on all dimensions of the Short Form-36 (SF-36) compared to healthy individuals. Around one-third of the patients suffered from anxiety and depression (Hospital Anxiety and Depression scale) (Ahn et al. 2010). Nonetheless, the indications for EA and religious, social and cultural backgrounds are different for patients in Korea and Western counties.

We have studied HRQOL, perceived stress, self-rated health, job separation because of illness or disability and socioeconomic position of EA patients in Denmark and compared these results with the general Danish population.

Material and Methods

Material

In 2005, the hospital database at Rigshospitalet, Copenhagen, was screened using surgery codes (ICD 10) for patients who had undergone bulbar evisceration, enucleation, orbital exenteration or secondary implantation of an orbital implant during the period 1996–2003. Rigshospitalet is a public, national specialized hospital with a recruitment area of eastern Denmark (approximately 2 million people). Nearly, all cases of EA in eastern Denmark were treated at Rigshospitalet in this period of time (Rasmussen et al. 2010).

Of the 431 patients identified from the database, 126 were deceased, seven were younger than 18 years, eight had left the country, and 23 patients were from Greenland or the Faeroe Islands and thus were excluded. In August 2005, the remaining 267 patients were invited to participate in a clinical investigation: 59 patients refused, 28 did not reply and seven did not show up for their appointments. The remaining 173 patients had individual interviews and clinical examinations in December 2005, including measurement of visual acuity in the remaining eye, position and configuration of eyelids and eye, motility and signs of exposure of the implant. Medical records were collected and reviewed for information about causative diagnosis (the disease process leading to the clinical condition that resulted in the EA) and the final indication for EA (Roed Roed Rasmussen et al. 2009). The study was approved by the regional ethics committee (Copenhagen and Frederiksberg; No. 01272128).

Contact procedure

By February 2008, eleven of the original patients in the study had died and three patients specified that they did not wish to receive a follow-up questionnaire. The questionnaires were sent by ordinary mail, with a cover letter and a prepaid return envelope, to the remaining 159 patients. Patients who did not return the questionnaires after 1 month received a reminder along with new copy of the questionnaire and another prepaid envelope. Some relatives responded on behalf of seven patients: the patients were senile, blind or severely ill and therefore not able to complete the questionnaires. In total, 131 patients completed and returned the questionnaire.

We defined patients with visual acuity of 6/36 or less as visually impaired (= 11). Data for this group were analysed separately because of this severe handicap. Of the total sample of 120 patients, 63 had evisceration, 55 enucleation and two exenteration of the orbit. Median observation time since EA was 5 years [range: 2–64 years].

Instruments used

The questionnaire included the following: Short Form-36 (SF-36), Perceived Stress Scale (PSS) and questions about self-rated health, job separation because of illness or disability as well as socioeconomic position.

The SF-36 is an instrument that measures HRQOL. It contains 36 items measuring eight dimensions of health and well-being: ‘physical functioning’ (PF), ‘role limitations due to physical problems’ (RP), ‘bodily pain’ (BP), ‘general health perceptions’ (GH), ‘vitality’ (VT), ‘social functioning’ (SF), ‘role limitations due to emotional problems’ (RE) and ‘mental health’ (MH). Each dimension is scored from 0 (worst possible health state) to 100 (best possible health state) (Bjorner et al. (1997)). The Danish adapted version was tested and validated in 1998 (Bjorner et al. 1998a,b,c). The self-administrated SF-36 has been used in various ophthalmological studies (Cruickshanks et al. 1999; Chia et al. 2003, 2004, 2006b; Varma et al. 2006; Elliott et al. 2009; Gall et al. 2009) and has been found useful for collecting data from older patients as well as visually impaired patients (Chia et al. 2006a).

The Perceived Stress Scale is a 10-item measure of the degree to which respondents appraise stressful situations that occurred during the past month (Cohen et al. 1983; Cole 1999; Sewitch et al. 2001). The items are scored on a 5-point scale, from 0 to 4, and the total score provides a global measurement of the extent to which an individual feels overwhelmed by stressful situations that occurred in the past month. Total scores range from 0 to 40; higher scores indicate greater perceived stress (Cohen et al. 1983; Cole 1999).

Questions about self-rated health, job separation because of illness or disability and socioeconomic position are taken from the Danish Health Interview Survey 2005 (Ekholm et al. 2009).

Self-rated health is measured in two ways. First, participants are asked to rate the following statement: ‘In general, would you say your health is: ‘(Excellent, very good, good, fair, poor)’. Second, the respondents are asked to indicate how they rate their own state of health that day, by drawing a line on a vertical 14-cm visual analogue scale (VAS), which quantifies their health: from 0 as the worst imaginable health to 100 as the best imaginable health. Self-rated health on a VAS is a part of the EuroQol EQ-5D scales (EQ-VAS). The EQ-5D scale has been used previously in Danish studies (Kessing et al. 2006; Vogel et al. 2006) and has been translated for and validated in many countries (http://www.euroqol.org/).

Job separation because of illness or disability were identified through the following two questions: (i) ‘Have you ever needed to change your working hours to part time, quit a job or change your job or work tasks due to disease, disorder, illness or injury?’ (Yes, worked part time; yes, changed job or work tasks; yes, stopped working; yes, first changed job or work tasks and then stopped working; no); (ii) ‘If the answer is yes to the previous question – was this because of eye disease?’ (Yes; no).

Marital status was determined by the question: ‘What is your marital status?’ Socioeconomic position was measured with one question about the subject’s occupation: ‘What is your occupation?’ Two questions were used to address leisure activities: (i) ‘Are there leisure activities that you no longer take part in due to the loss of your eye?’ (Yes; no) and (ii) ‘If yes, please write down the types of activities’.

Comparison group

Data from the EA patients were compared with data from the nationally representative Danish Health Interview Survey 2005 (Ekholm et al. 2009). The Danish National Institute of Public Health, University of Southern Denmark, has carried out national representative health interview surveys among adult Danes since 1987. The survey in 2005 was based on a region-stratified random sample of 21 832 Danish citizens aged 16 years or older. Data were collected via face-to-face interviews at the respondents’ homes (response rate: 67%). Following the interviews, all respondents were asked to complete a self-administered questionnaire. Information about sociodemographic characteristics and job changes because of illness or disability were collected via face-to-face interviews and the SF-36, the PSS and self-rated health on a VAS were included in the self-administered questionnaires. Because SF-36 and PSS were part of the Danish Health Interview Survey, we had to choose the same questionnaires for HRQOL and stress to compare with the general Danish population.

Statistical analyses

Because the standard errors for the general population estimates were very small, the estimates were considered to be ‘true’ values representing the adult Danish population. Thus, because the t-test do not require any assumption of normal distribution in adequately large samples (Lumley et al. 2002), the one-sample t-test was used to compare the study populations’ SF-36 and PSS mean scores with age-standardized mean values for the general Danish population (Danish normative data). Effect sizes (Cohen’s d) were also calculated as an indication of the magnitude of the mean differences between the study population and the general population. The effect sizes were calculated by dividing the difference between the means scores of each group by the pooled standard deviations for those means. Following Cohen’s recommendation, effect sizes of 0.20, 0.50 and 0.80 were considered to be small, moderate and large, respectively (Cohen 1988).

The nonparametric (Kruskal–Wallis) test was applied to compare the SF-36 and the PSS median scores between different subgroups in the study population. Furthermore, the one-proportion z-test was used to compare prevalences in the study population with age-adjusted prevalences from the general population.

Results

The sociodemographic characteristics of the study population and the general population are presented in Table 1. Approximately twice as many members of the sample of patients (Table 1) are divorced or separated compared to the general population (12.5% versus 6.3%).

Table 1.   Sociodemographic characteristics of the study population and of the Danish Health Interview Survey 2005. Percentage.
 Study populationGeneral population
Sex
 Men50.048.4
 Women50.051.6
Age
 <44 year41.743.7
 44–64 year31.735.9
 >65 year26.720.4
Marital status
 Married48.354.4
 Cohabiting15.015.5
 Single (separated, divorced)12.56.3
 Single (widowed)8.36.9
 Single (unmarried)14.216.9
 No information1.70.0
 No. of respondents12014 566

Socioeconomic position of the EA patients was categorized as follows: employed (= 53), self-employed (= 8), unemployed (= 4), student (= 2), retired pensioner (= 35) and disability pensioner (= 12). Six participants had missing values for socioeconomic position.

In the study population, 39.5% stated there were leisure activities they no longer took part in because of their EAs. The types of leisure activities the patients had to discontinue included the following: ball games (= 20); physical activities like walking, hiking, running and aerobics (= 7); bicycling (= 7); swimming (= 7); reading books and other activities requiring concentration (= 6); car driving (= 4); social activities (= 2); household chores (= 3); and shooting, skiing, travelling and cultural contexts (= 7).

The EA patients had significantly (p < 0.05) lower scores (poorer health) for all eight SF-36 subscales and more perceived stress compared to the general population (Table 2). Effect sizes indicated that the largest differences in HRQOL were related to role limitations because of emotional problems (= −0.47) and mental health (= −0.46). Furthermore, the analyses indicated that visually impaired patients report poorer health and health-related quality of life i.e. lower median scores on all eight SF-36 dimensions of health than other EA patients (data not shown).

Table 2.   SF-36 and PSS mean scores in the study population and in the general population.
 Study population (all indications) Mean (SD)General population (age-standardized) Meanp valueEffect sizea
  1. a Effect sizes of 0.20, 0.50 and 0.80 were considered to be small, moderate and large, respectively.

Bodily Pain (BP)70.2 (29.9)78.10.0047−0.33
General Health (GH)69.0 (26.3)75.40.0095−0.32
Mental Health (MH)75.9 (21.2)83.00.0004−0.46
Physical Functioning (PF)82.5 (24.4)87.60.0251−0.26
Role Emotional (RE)72.7 (38.4)86.80.0001−0.47
Role Physical (RP)67.2 (41.4)81.30.0004−0.41
Social Functioning (SF)84.7 (24.1)91.90.0014−0.40
Vitality (VT)62.8 (23.4)69.90.0012−0.34
PSS12.4 (6.9)11.00.03970.23
No. of respondents12011 238  
Mean age (years)51.4 (17.0)48.9  

In Table 3, the SF-36 score and PSS median are stratified on the youngest (≤44 years) and the eldest (≥45). The analysis shows a significant difference in the subscale physical functioning (p = 0.0036).

Table 3.   SF-36 and PSS median scores (range) by age.
 ≤44 y.≥45p value
Bodily Pain (BP)84 (100)74 (100)0.5239
General Health (GH)83.5 (100)72 (95)0.0500
Mental Health (MH)84 (100)84 (88)0.2745
Physical Functioning (PF)100 (100)90 (100)0.0036
Role Emotional (RE)100 (100)100 (100)0.2569
Role Physical (RP)100 (100)75 (100)0.0720
Social Functioning (SF)100 (100)100 (62)0.9000
Vitality (VT)65 (95)65 (95)0.7821
PSS10.5 (39)11 (27)0.8458
No. of respondents5070 
Mean age (years)35.163.0 

The SF-36 scores and PSS median, stratified on the four main indications for EA, are listed in Table 4. The analysis did not show a statistically significant difference in median scores (all p > 0.05) between the four main indications for EA. However, the analysis indicated that painful blind eye, as an indication for EA, had the largest impact on health-related quality of life and more perceived stress. It is noteworthy that the mean age in this group is 50.6 years. Other indications for EA were as follows: infection (= 8), prevention of sympathetic ophthalmia (= 3) and unknown reasons (= 2). Because of their rarity, data from patients with these indications were not analysed separately.

Table 4.   SF-36 and PSS median scores (range) by indication for eye amputation.
 Painful blind eyeNeoplasmTraumaDisfiguring blind eyep value
Bodily Pain (BP)62 (100)84 (78)84 (100)62 (100)0.2010
General Health (GH)62 (95)82 (75)77 (100)69.5 (90)0.2541
Mental Health (MH)76 (100)84 (68)84 (80)80 (80)0.0765
Physical Functioning (PF)90 (100)95 (60)95 (67)90 (80)0.2451
Role Emotional (RE)83.5 (100)100 (100)100 (100)100 (100)0.5121
Role Physical (RP)100 (100)87.5 (100)100 (100)100 (100)0.9544
Social Functioning (SF)88 (100)100 (75)100 (87)100 (87)0.0548
Vitality (VT)55 (90)70 (90)65 (80)60 (78)0.1182
PSS14.5 (39)10 (30)11.5 (27)9 (24)0.4145
No. of respondents38302115 
Mean age (years)50.657.842.252.3 

In Table 5, the eight dimensions of SF-36 and the PSS are stratified for the presence of phantom eye syndrome, observation time over years (2–9 years and more than 9 years) and type of surgery used for treatment.

Table 5.   SF-36 and PSS median scores (range) in different diagnostic subgroups.
 Phantom Eye SyndromeObservation time in yearsType of surgery for EA
YesNop value2–9 year≥10 yearp valueEviscerationEnucleationp value
Bodily Pain (BP)62 (100)84 (100)0.005180 (100)82 (78)0.408067 (100)84 (88)0.1928
General Health (GH)72 (100)77 (90)0.182472 (100)77 (80)0.794972 (100)77 (80)0.3801
Mental Health (MH)84 (88)84 (100)0.185484 (100)82 (72)0.802780 (100)88 (72)0.0128
Physical Functioning (PF)95 (100)90 (100)0.993595 (100)100 (67)0.044390 (100)95 (67)0.1360
Role Emotional (RE)100 (100)100 (100)0.2649100 (100)100 (100)0.4346100 (100)100 (100)0.3705
Role Physical (RP)100 (100)100 (100)0.2165100 (100)100 (100)0.3689100 (100)100 (100)0.5726
Social Functioning (SF)100 (87)100 (100)0.0883100 (100)94 (87)0.1293100 (100)100 (87)0.0502
Vitality (VT)65 (95)67.5 (100)0.130765 (100)65 (68)0.535260 (90)70 (90)0.0682
PSS12.5 (30)10 (39)0.659910 (39)12 (29)0.316212 (39)10.5 (30)0.7882
No. of respondents6159 8832 6355 
Mean age (years)50.951.9 53.346.2 49.453.5 

Patients with phantom eye syndrome had a significantly lower median score (p = 0.0051) on the dimension for bodily pain, than did patients without phantom eye syndrome. The type of surgery used for treatment seemed to have only a minor impact, because among the eight dimensions of the SF-36 and the PSS, only the dimension of mental health (p = 0.0128) was more frequently associated with lower median values in evisceration than enucleation (Table 5). Two patients received an orbital exenteration because of a neoplasm. No specific analyses were carried out for this small group. Both patients received an exoprosthesis.

In all, 43.3% of the EA patients had excellent or very good self-rated health (Table 6). This percentage is lower than for the general population (52.1%). However, the difference was not statistically significant (p = 0.05). Measuring self-rated health by means of the VAS also indicated (p < 0.01) that patients experience poorer health than does the general population (study sample: mean: 70.4, SD: 22.2; general population: age-standardized mean: 78.6, SD: 17.2). The visually impaired group had a mean VAS score of 59.5 (SD: 18.9).

Table 6.   Comparison of self-rated health and job separation because of illness or disability between the study population and the general Danish population.
 Study population (Raw % and 95% CI)General population (Age-standardized %)p value
Excellent or very good self-rated health43.3 (34.5–52.2)52.10.0546
Changed job because of illness20.0 (12.8–27.2)7.4<0.0001
Retired to due illness25.8 (18.0–33.7)9.2<0.0001
Part-time because of illness10.8 (5.3–16.4)4.10.0002
No. of respondents12014 566 

Thirty-one participants in the study population were retired because of illness; of those, 18 (58%) said it was because of their eye disease. Thirteen patients only worked part-time; eight (62%) of those because of the eye disease. A job change because of illness was noted by 24 patients, out of which 13 (54%) reported that it was because of the eye disease. In total, 25% of the study population retired or changed to part-time jobs because of their eye diseases (Table 6).

Discussion

Our study indicates that EA patients have poorer HRQOL and self-rated health and more perceived stress than the general population does. Recovery after loss of an eye has been investigated in two previous questionnaire studies (Linberg et al. 1988; Coday et al. 2002). In 1988, one study found that among 125 EA patients, 37% had experienced permanent changes in their lives because of the EA, 10% had problems with employment and 17% suffered from anxiety or poor self-image (Linberg et al. 1988). In 2002, a questionnaire investigation of 58 EA patients and seven functionally one-eyed patients found changes in employment (23%), changes in driving status (39%), socially affected lives (40%) and difficulties with sports and hobbies (50%) (Coday et al. 2002). These observations are similar to our findings, despite that patients in both prior studies had an average age around 25 years and 60–67% of the EAs were performed because of a recent trauma (Linberg et al. 1988; Coday et al. 2002). In the present study, the average age of EA patients is 51.4 years, and the most common indications were painful blind eye and neoplasm.

Health-related quality of life among 134 anophthalmic patients in Korea was investigated in 2008 and compared to 48 healthy volunteers. We presume that the authors defined anophthalmic to mean patients who have lost their eyes and not patients who were born anophthalmic/micro-ophthalmic. The anophthalmic Korean patients scored significantly lower on all eight domains of the SF-36 compared to the healthy volunteers. However, many differences exist between the Korean and Danish study populations concerning cultural, economic, social and religious factors. Furthermore, the sociodemographic factors are diverse: Koreans in the sample were around 10 years younger than were the Danes, 10% more of the Korean sample participants were married, and over 50% had lost their eyes because of a trauma. It is interesting there are similar results concerning SF-36 analyses in comparison with healthy individuals, as in our sample comparing patients with the general Danish population (Ahn et al. 2010). We do not have access to the raw SF-36 scores from the Korean study, but it appears that the levels on all eight domains are lower for the healthy Korean volunteers than for the general Danish population (see Table 2). Data from the Korean sample also indicated that one-third of the patients exhibited anxiety and depression (Hospital Anxiety and Depression scale).

Differences in HRQOL scores of mono- and bilaterally visually impaired patients have also been demonstrated using the SF-36. Patients with monolateral visual impairment had lower scores on three dimensions, and bilateral patients had lower scores on five dimensions compared to persons with no visual impairment (Chia et al. 2003, 2004). Our study supports this. The eleven patients, visually impaired in the only remaining eye, had lower SF-36 scores in all eight dimensions.

Younger EA patients (44 years or younger) had a significantly higher score on the SF-36 PF scale than the older EA patients (45 years or older) indicating a better physical function in the youngest age group. Reduced physical function with increasing age is also seen in the general population (Bjørner et al. 1997).

Patients with painful blind eye and disfiguring blind eye as indication for EA reported, although not significant, poorer HRQOL scores and more perceived stress than did patients with cancer and trauma (Table 4).

For patients with cancer, the EA was part of a life-saving treatment. Studies of the relationship between cancer and HRQOL have found that survivors of cancer have almost the same levels of HRQOL as the general population, which is similar to the findings from our study (Eiser et al. 2000; Peuckmann et al. 2007). Such findings could be interpreted as changes in the patients’ internal standards, values and conceptualizations of life over time because of survival of cancer: a ‘response shift’ (Sprangers 2002). Studies of facially disfigured patients have indicated that the course of injury is related to self-consciousness and anxiety, which supports our findings (Tebble et al. 2004, 2006). Patients with trauma also tend to have a good HRQOL similar to the patients with cancer. Like the patients with cancer, patients with trauma as indication for EA did loose their eye within the first 14 days after the initial trauma, and in most cases the surgery was inevitable. Type of surgery indicated the same tendency, i.e. eviscerated patients having lower median values in most dimensions of SF-36, and more perceived stress. Eviscerated patients generally have the indication painful or disfiguring blind eye, and all patients with cancer are enucleated (Table 5) (Rasmussen et al. 2010). Studies of patients with disfigurements have found that levels of psychosocial distress are not well predicted by the severity of disfigurement (Dropkin 2001; Rumsey et al. 2003).

Effect sizes indicated that the largest differences in HRQOL between the EA patients and the general population were related to role limitations because of emotional problems and mental health. One might speculate that feelings of being different and looking different (altered body images) result in poor self-confidence and self-image. This is also demonstrated by the lower self-rated health among the EA patients, as seen in Table 6.

We found no differences between patients with short- and long-observation time (Table 5) with the exception of physical functioning. The presence of phantom eye syndrome did not predict significantly poorer HRQOL or more perceived stress, except for the dimension of bodily pain (Table 5), which is to be expected, because around 44% of patients with phantom eye syndrome indicate phantom pain (Roed Roed Rasmussen et al. 2009; Soros et al. 2003, 2005). In the current sample of patients, the incidence of visual hallucinations was 42% and phantom pain was 23% (Roed Roed Rasmussen et al. 2009). From the interviews, we knew that many of the patients find the visual hallucinations disturbing, and most patients keep their visual hallucinations a secret (Roed Roed Rasmussen et al. 2009).

Strengths and limitations

Information about job separation because of illness or disability among the general population sample was obtained via face-to-face interviews, and this may have influenced the comparisons between the study population and the general population. However, the different modes of data collection cannot explain the vast differences in prevalence of job separation because of illness or disability between the two populations.

The time span between the measured visual acuity and administration of the questionnaires was two years (from December 2005–February 2008). In this time span, visual acuity may have worsened. But again, the majority of the patients had good visual acuity in the remaining eye and combined with the median age of 50 years, we do not think this can explain the results.

In conclusion, we found that EA patients have poorer HRQOL, poorer self-rated health and more perceived stress than does the general population. Effect sizes showed that the largest differences in HRQOL between the EA patients and the general population were related to role limitations because of emotional problems and mental health. Furthermore, 25% retired or changed to part-time jobs because of eye disease and 39.5% stopped participating in leisure activities because of their EAs.

Acknowledgements

This work was financially supported by the Velux Foundation, Soeborg, Denmark. Grant number: 270479. There are no conflicts of interest. Furthermore, this part of the study has not been possible without inspiring support from Finn Kamper-Jørgensen and the cooperation with the National Institute of Public Health, University of Southern Denmark.

Ancillary