Article Text

Download PDFPDF

Reduced mortality compared with national averages following phacoemulsification cataract surgery: a retrospective observational study
  1. M S J Blundell1,2,
  2. L P Hunt3,
  3. E J Mayer1,2,
  4. A D Dick1,2,
  5. J M Sparrow2,4
  1. 1
    Academic Unit of Ophthalmology, University of Bristol, Bristol, UK
  2. 2
    Bristol Eye Hospital, University Hospitals of Bristol NHS Trust, Bristol, UK
  3. 3
    Department of Clinical Sciences South Bristol, University of Bristol, Bristol, UK
  4. 4
    International Centre for Eye Health, London School of Hygiene & Tropical Medicine, London, UK
  1. Mr M S J Blundell, c/o Clinical Research Unit, Bristol Eye Hospital, Lower Maudlin Street, Bristol BS1 2LX, UK; msjb1{at}


Background: Higher or equal rates of mortality are associated with cataract surgery compared with the general population. Cataract surgery has advanced, and the clinical characteristics of the patient undergoing cataract surgery have changed.

Aims: To reinvestigate survival following cataract surgery.

Method: Survival data were gathered up to the end of 2006 on 933 consecutive patients who underwent cataract surgery between December 2000 and February 2001. These data were compared with national and regional mortality figures, and standardised mortality ratios (SMR) were calculated.

Results: After adjusting for age and sex, there was a statistically significant reduced mortality compared with national (SMR = 0.88 (95% CI 0.79 to 0.99)) and regional figures (SMR = 0.87 (95% CI 0.78 to 0.98)).

Conclusion: All previous studies found decreased survival among cataract surgery cohorts. These data differ from data at earlier times, as cataract surgery seems to be associated with increased survival. This illustrates the need for continual re-evaluation of accepted medical knowledge in the light of changes in practice and population demographics.

Statistics from

A large number of studies have investigated mortality associated with cataract 115 (see table 1) or following cataract surgery 10 12 13 1622 (see table 2).

Table 1 Studies investigating mortality associated with presence of cataract on clinical examination
Table 2 Studies investigating mortality following cataract extraction

It is useful to consider those who undergo surgery because this is the group with whom ophthalmologists have the most clinical contact. All of the statistically significant results from studies assessing survival in this group have shown an increased mortality following cataract surgery. However, most of these studies look at patients who underwent cataract extraction prior to 1992. Since this time, there have been significant changes to the process of cataract surgery and the clinical characteristics of the patient group who undergo cataract surgery.23


We estimated a 3-month period within our single unit would yield 1000 patients for analysis. All consecutive patients who underwent cataract surgery at the Bristol Eye Hospital between 1 December 2000 and 1 March 2001 therefore were included.

Date of birth, date of operation and gender were obtained from the hospital database. The National Health Service Strategic Tracing Service (NSTS) was used to obtain the date of death. The NSTS enabled us to include patients who had moved out of the region, but had no information for patients who left the UK. Deaths that had occurred up to and including 31 December 2006 were included. Survival status was assessed on 11 July 2007 to allow for the delay in the registering of deaths.

The national statistics agency provided age- and gender-specific population totals as well as number of deaths from all causes per year from 2000 to 2006, for both England and Wales, and the Bristol region. These were used to calculate national and regional age- and gender-specific mortalities. We used both, since, while the rates for England and Wales were estimated on larger numbers, and were therefore more precise, regional variation existed. These rates were used to calculate the expected numbers of deaths for the cataract patient cohort, using a method based on the Lexis diagram.24 The observed numbers of deaths (O) was compared with the expected number (E) by calculating the Standardised Mortality Ratio (SMR) = O/E, together with its approximate 95% CI.25

The patient cohort included a small number of young patients with cataract. As these patients would be expected to have congenital cataract or secondary cataract, it is likely that they represent a group with different clinical characteristics than those who had senile cataract. Analysis was performed on the total group and on the group who were 50 years and above at the time of surgery to address this issue.


A total of 933 patients underwent cataract surgery from December 2000 to February 2001. Demographic data for these patients are given in table 3.

Table 3 Baseline characteristics of the study population

The median age at surgery was 77.8 years (range 6.1 to 98.7 years). One-third of patients were men.

A total of 298 patients had died by the end of the study follow-up period. A breakdown of deaths by calendar year is given in table 4.

Table 4 Mortality results: all patients (n = 933)

There were fewer deaths among the cataract surgery cohort in the period from earlier years after surgery (up to 2005) than expected, using either the national rates or regional rates, but these differences were not statistically significant. The mortality for the whole period, 2000 to 2006, was significantly reduced with respect to either the national (SMR = 0.88 (95% CI 0.79 to 0.99)) or regional figures (SMR = 0.87 (95% CI 0.78 to 0.98)). After removing patients under 50 years at the time of surgery, in an attempt to remove non-age-related cataracts, the overall SMR was unchanged: 0.88 (95% CI 0.78 to 0.98) compared with the national population and 0.87 (95% CI 0.78 to 0.97) compared with the regional population.

A breakdown of SMR by age group is shown in table 5 using national rates.

Table 5 Mortality results: breakdown by age at operation

Mortalities were higher than expected for the under 70s and lower than expected in those aged 80+. Only the 60–69 and 80–89 year age groups achieved statistical significance (SMR = 2.28 (95% CI 1.56 to 3.32) and 0.72 (95% CI 0.72 to 0.85), respectively).


Many articles show increased mortality among those with cataract.110 1315 A number have shown that mortality is increased in diabetics with cataracts after adjusting for age and gender.13 The authors suggested that cataract in patients with diabetes is an indicator of poor glucose control, therefore associated with systemic complications and mortality. Other studies found increased mortality associated with cataract independent of diabetes.4 5 7 911 14 15 Many of these authors put forward the hypothesis that cataract is a sign of frailty, or of general tissue senescence and comorbid systemic ill health in an attempt to explain this association.

A number of articles have shown increased mortality among patients undergoing cataract surgery.12 13 1622 A study by McGwin et al13 showed an increased overall mortality in cataract patients compared with non cataract patients, but among the cataract patients the increased mortality was less in the group that underwent cataract surgery. This difference remained after multivariate adjustment for other causes of mortality. A study by Wang et al also highlighted this difference showing an increased mortality in patients with cataract but a decreased mortality in those who had undergone cataract surgery. Their results, however, failed to achieve statistical significance.

If the only reason for increased mortality in cataract patients is that cataract is a measure of general frailty, cataract surgery should not affect survival. The difference in mortality between those who have not had cataract surgery and those who have suggests that other mechanisms may be involved.

Our study provides evidence towards the hypothesis that there is no increased mortality among patients who have had cataract operations compared with the national population.

This agrees with evidence presented in Wang et al’s paper.10 It is difficult to make comparisons between the McGwin study and our own, as the McGwin study compared a cataract surgery group to a group without cataract but with another ophthalmic diagnosis. Our study compared a cataract surgery group to the national population. All previous studies that have examined mortality following cataract surgery compared with a national population, which can therefore be compared directly to our study, found decreased survival among the cataract surgery cohort.1821 In order to provide an explanation for this discrepancy, let us consider differences between the cohorts. The most striking difference in these previous studies is that cataract surgery took place prior to 1992, while our cohort had surgery in 2000/2001.

Over the last 15–20 years, cataract surgery trends have changed dramatically. New technology has changed the surgical technique from intracapsular to extracapsular to phacoemulsification cataract surgery in a relatively small time span. This change in technique has been associated with patients being operated on with better preoperative vision23 and, therefore, probably with less cataract. Surgeons are now operating on greater numbers of patients over a greater age range. Overall a greater proportion of the population is undergoing surgery.23 The current findings suggest that changes in the characteristics of those undergoing surgery may have shifted the cataract surgical population closer to the general population with consequently more similar patterns of survival.

Age is a key risk factor for mortality. The average age at surgery in our study was 77.8 years. The average age was given in three18 20 21 of the four studies that used a national population as a reference population and was slightly younger, ranging between 71 and 75.3 years.

Subanalysis of SMR by age group in our study showed a decreasing SMR with increasing age (see table 5). All of the other four studies also performed this subanalysis and showed a decrease in relative risk, hazard ratio or SMR1821 with increasing age. Our study showed the crossover from an increased SMR of 2.28 (95% CI 1.56 to 3.32) in the age group 60–69 years to an SMR close to the national population (1.02, 95% CI 0.84 to 1.25) in the age group 70–79 years to a decreased SMR of 0.72 (95% CI 0.61 to 0.85) for the 80–89 year age group. Although the average age at surgery in our study is only slightly higher than in the other comparable studies, it sits in an age group that represents a crossover point when looking at relative mortality. This could partially explain the difference in mortality between the studies.

Our study used mortality data collected from the NSTS; this would not include data from people that had left the UK. The patients who we have dates of death for we know died in the UK. Those who we do not have dates of death for we presumed were alive at the end of the 6-year follow-up, but it was possible that they had left the UK and died abroad. In an attempt to account for patients in our cohort that had moved out of the UK, we used estimated migration rates from the south-west of England, as provided to us by the National Statistics Office for five age-bands over the periods 2001–3 and 2004–6. We were able to calculate that approximately eight of the 635 survivors may have emigrated. If we were to assume the worst case scenario that these eight had died within the period up to the end of 2006, the overall SMR would still be less than 1 (SMR = 306/338.670 = 0.90; 95% CI 0.81 to 1.01).

There are other possible reasons to explain the difference in results between our study and those that have examined this subject previously. Patients with better vision as a result of having undergone cataract surgery may be less likely to injure themselves and might therefore have a lower mortality. Wang et al examined causes of death for pseudophakic patients compared with patients with cataracts and found no increase in deaths caused by injuries or fractures, suggesting that this is not the reason for the difference in mortality.

By only considering patients who have undergone cataract surgery, it is likely that we are introducing an element of bias. The fact that cataract surgery is available and is free in the UK NHS system at the point of delivery to UK residents should negate socio-economic differences. However, cost and availability are not the only reasons for failure to access medical care. A study by Keenan et al26 shows wide variation in rates of cataract surgery by local authority across the UK, but was unable to distinguish whether such variations were explained by differences in prevalence or by referral patterns. Similarly, in our study we cannot exclude a difference between the populations we have compared in their ability to access medical care that could potentially explain the reduced mortality.

Through only selecting patients who are healthy enough to undergo cataract surgery we could be selecting a healthier population than the national population. However, as there are only a small number of comorbidities that would directly prevent cataract surgery being performed, and as the vast majority of cataract operations are carried out under local anaesthetic, few would be considered unfit for surgery. Despite this, we cannot exclude the fact that some patients may not undergo cataract surgery due to chronic ill health.

A limitation of the study is that data were only collected from one hospital yet compared with the national population. This has a number of implications.

Regional variations exist for both mortalities and for rates of cataract surgery.26 We have tried to address this by using both national and regional mortality figures to compare with our cohort. The fact that the overall result remains the same is a good indication that these data may be representative for the nation.

As the population who had cataract extraction was compared with the overall population, not with the population who had not had cataract extraction, the result is an underestimate of the true difference in mortality between the groups.

The study was retrospective, so we were not able to collect data on comorbidity at time of surgery. Subsequently we were not able to adjust for disease, for example diabetes. If the overall level of disease in the general population was greater than in the study population, this could account for the reduced rate of mortality. However, other studies have noted that adjusting for systemic factors led to a reduced hazard ratio associated with cataract.5 6 15 This indicates a higher proportion of patients with disease among a population with cataract, so we would expect a lower observed mortality if we could adjust for disease.

The advantages of our study are the large patient cohort (n = 933) compared with national and regional population totals, and the relatively long follow-up of 6 years.

In summary, our study gives evidence towards the hypothesis that there is no increased mortality following cataract surgery. This is a subject that has had little attention since the dramatic changes that have occurred in the technique and delivery of cataract surgery over the last 15 years, which have led to a great increase in the volume of cataract surgery performed. Our study finds a different result from the majority of previous survival studies undertaken on patients undergoing cataract surgery at an earlier time. This study illustrates the need for continual re-evaluation of seemingly accepted medical knowledge in the light of changes in practice and population demographics.


The authors would like to acknowledge the Office of National Statistics for providing national and regional comparison data.



  • Funding: MB is funded as a Clinical Research Fellow by Bausch and Lomb. EM is funded by a National Career Scientist Award from the Department of Health and NHS Research and Development.

  • Competing interests: None.

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.

Linked Articles

  • Editorial
    Syril Dorairaj Kannan M Narayana Parasappa Bandrakalli