- Top of page
Objectives: To study non-tolerance to spectacle prescriptions in a busy community optometric practice, with several practitioners.
Methods: A spectacle non-tolerance was defined in a pragmatic way, as a patient who had collected spectacles from the practice and subsequently returned because they were either having problems with, or were unable to wear, their new spectacles. Patients over 16 years of age, who met the above definition of non-tolerance were sequentially recruited over a 6 month period. Patients experiencing adaptation problems were first seen by a dispensing optician and any dispensing issues resolved. If the spectacle dispensing was felt to be correct, or if the non-tolerance persisted, then the patient was re-examined by an optometrist and the results analysed.
Results: Non-tolerance examinations accounted for 62 of the 3091 eye examinations during the study period. The average rate of non-tolerance, averaged across the practitioners, was 1.8%, varying from 1.3–3.3% for individual practitioners. Gender was not a factor in non-tolerance, however age was, with presbyopes accounting for 88.1%. The most common reasons for non-tolerance were, in order of decreasing frequency: prescription related (61.0%), dispensing related (22.0%), pathology (8.5%), data entry error (6.8%) and binocular vision anomalies (1.7%). Of prescription related errors, gauging the spherical element accounted for the majority of inaccuracies, followed by problems with the near/intermediate addition. In every case, the final prescription was within 1.00 D of the not tolerated, prescription; 84.4% were within ±0.50 D.
Conclusions: Spectacle prescription non-tolerance forms a small, but important, form of adverse reaction in optometric clinics. Most non-tolerances can be resolved by small changes, within 0.50 D, to the prescription.
- Top of page
In recent years, there has been an increased emphasis in the healthcare services on clinical governance and on causes of adverse reactions. It is perhaps surprising therefore that there have been relatively few studies into adverse reactions to optical prescriptions, commonly known as ‘non-tolerances’. This study was an evaluation of non-tolerance cases in a large independent optometric practice. The main aim was to determine the most common reasons for a patient to return unsatisfied with their new spectacles. Additionally, the normal or average rate for returning with those spectacles was calculated and comparisons were made between the non-tolerance rate for different practitioners.
Firstly what do we mean by non-tolerance? Non-tolerance to spectacles can be divided into two categories (Priest, 1979):
Pubmed searches for keywords such as spectacle/glasses prescription errors, prescribing spectacles, reliability of refraction and dissatisfied refraction patient revealed that there was limited information on this subject. Some searches, for example, spectacle/glasses non-tolerance produced no information. Nonetheless, the optometric literature does provide information on common causes for spectacle prescription non-tolerance. These are listed in Table 1 (Ball, 1977). There are many symptoms that a patient can experience following a new optical prescription. Some of these will be temporary and settle down, others will be more persistent (see Table 2 (Ball, 1977)).
Table 1. Major causes of non-tolerance to optical prescriptions. (Table reproduced with kind permission of Optician: First published Oct 14, 1977.)
|Practitioner orientated||Dispensing errors and associated problems|
|Faulty refraction and prescription|
|Undetected or subsequently developed abnormality|
|Management of initial examination|
|Patient orientated||Adaptation problems|
|Motivation; expectation; dissatisfaction|
|Practitioner/patient relationship||Attitudes; personality patterns|
Table 2. Transient or persistent adverse effects of optical prescriptions. Modified after Ball (1977)
|Symptom-type||Adverse effects or presenting symptom||Examples of possible aetiology|
|Sympathetic or referred||Headaches and/or discomfort referred to ocular adexna||High presbyopic additions|
|Faulty lens centration|
|Mild vertigo or dizziness||New prismatic corrections|
|Relative prismatic effects|
|Disorders of visual perception and binocular vision||Micropsia||Recently corrected myopia|
|Base out prisms|
|Macropsia||Recent presbyopic additions|
|Base in prisms|
|General spatial distortions||Corrected anisometropia|
|Change of lens form|
|Peripheral spatial distortion||Some multifocals|
|Blurred vision||Incorrect effective power|
|Incorrect positioning of bifocals/multifocals|
|Off axis blur from high index lenses|
|Inappropriate vocational use|
|Residual uncorrected errors|
|Diplopia||Faulty lens centration|
|Others||Chromatopsia||High additions in some fused bifocals|
|Photophobia||Omission of previously worn tints|
|Ghost images||Reflection from lens or other surfaces|
The aim of subjective refraction is ‘to provide the patient with the optical correction nearest to the optical ideal with which he sees best and is most comfortable’ (Duke-Elder and Abrams, 1970). In optometric practice, patients are prescribed lenses to the nearest 0.25 D (Freeman and Hodd, 1955; Jalie, 1988; Miller et al., 1997; Smith, 2006). Goss and Grosvenor reviewed literature on several studies about reliability of refraction and concluded that conventional refraction is repeatable to within 0.25 D in approximately 80% of cases and 0.50 D in 95% of cases (Goss and Grosvenor, 1996; Smith, 2006). A study using multiple community optometrists found that the 95% reproducibility limits for stigmatic data was 0.78 D (MacKenzie, 2008). This agrees with research used standardised patient methodology (Shah et al., 2007) which found that subjective refractive findings are reproducible to approximately ±0.75 D when performed by multiple optometrists in patients of different age groups and levels of ametropia (Shah et al., 2009).
Small focal errors however can have an impact on critical tasks and for critical patients (Miller et al., 1997). Miller et al.’s study showed that a significant number of wearers notice errors in distance vision, as small as +0.25 D in sphere and cylinder. Miller et al. studied binocular errors only. In practice, however, asymmetric errors will also occur, affecting binocular co-ordination (Atchison et al., 2001) and sometimes producing asthenopia. A difference in refraction of 0.25 D between the two eyes causes a difference in size between the two retinal images of approximately 0.5% (Lowestein, cited in Duke-Elder and Abrams (1970)Comas et al., 2007). The limit that can normally be tolerated by most patients is 5% (Lowestein, cited in Duke-Elder and Abrams (1970)). Lovasik and Szymkiw (1985) showed that stereoacuity could be maintained to a clinically acceptable level of 40 arc seconds, using the Titmus circles test with a 0.50 D monocular defocus. They also found that stereopsis could be maintained in most subjects with a 2 D monocular blur, with 20% maintaining gross stereopsis with a 4 D blur.
More recent research by Atchison et al. (2001) investigated the effect of small prescription errors on spatial visual performance and spectacle lens acceptability. They collected data on 15 young adults and measured visual performance with best correction and for ±0.50 D spherical binocular and monocular errors. The results showed that prescription errors had a minimal effect on positive relative accommodation, distance and near heterophoria, and stereopsis. Prescription errors did however have significant effects on binocular visual acuity and negative relative accommodation.
An older study in 1971 by Appleton conflicts with the above theories. Appleton states that patients are not sensitive to small prescription changes and that rounding prescriptions off to the nearest half dioptre would not result in a significant increase in non-tolerance (Appleton, 1971). The study surveyed 91 patients who either had no change to the final prescription, or had a change of −0.25 D sphere, or −0.25 D cylinder. There was no statistical difference in dissatisfaction of the three groups. Looking at the bigger picture, however, there was a high dissatisfaction rate of 40%, among those whose lenses were not changed. In addition, the majority of patients in the study were pre-presbyopic, and therefore over minussing in this age group is more likely to be tolerated (Atchison et al., 2001).
Our literature search found little information available concerning the average rate for return of spectacles. Riffenburgh et al., (1983) reported from a non-surgical ophthalmological practice that 2.3% of 5467 patients were dissatisfied with their new spectacles. In a later study by Mwanza and Kabasele (1998), 2.8% of 432 patients returned with problems with their spectacles. A recent retrospective research study by Hrynchak evaluated a random sample of 25 718 records from a University optometric clinic. The rate of return was found to be 1.6% (Hrynchak, 2006). Another recent study (Steele et al., 2006) did a risk assessment analysis to determine errors/untoward events found in community optometric practices and found optical prescription errors accounted for 18.2%. This, however, included errors in dispensing spectacles.
We were unable to find any prospective studies involving community optometric practices that have sought to determine the rate of return of spectacles and also the average rate of return per practitioner. The present research was designed to provide this information and to establish the main reasons for return of spectacles, with its implications for primary eyecare practitioners. The Department of Health, (cited in Steele et al. (2006)), highlights the importance of finding and recording errors as a means of learning from experience.
- Top of page
The aim of this study was to investigate prescription non-tolerance in a busy community optometric practice. The protocol enabled the study to occur concurrently with routine practice work. Patients experiencing difficulties with their new (or reglazed) spectacles were seen first by a dispensing optician at the practice and an appointment was only arranged with an optometrist if the patient’s problems could not be solved in the first instance by the dispensing optician. This is in line with usual practice in the UK (Constantine-Smith, 2002) and the present research only investigated the cases that could not be satisfactorily resolved by the dispensing optician. This means that trivial problems (e.g. patients unhappy with their choice of frame colour or needing a simple frame adjustment) were excluded from the research. But it is also likely that some more serious errors (e.g. spectacles dispensed of an inappropriate prescription or lens type) might have been corrected outside of the study. Hence, the results are likely to have under-estimated the true prevalence of dispensing errors and data entry errors. Several other studies have followed a similar procedure, only studying cases where the non-tolerance required the attention of the prescribing clinician (Riffenburgh et al., 1983; Mwanza and Kabasele, 1998; Hrynchak, 2006). A minority of studies have looked at adverse reactions to new spectacles occurring at any stage in community optical practices (Steele et al., 2006) and this would be an interesting area for future research.
According to Strang et al. (1998), the average time taken to adapt to spectacles is 1 week. Fifty-six of the patients in the present study returned after one week, with only three patients reporting that they could not tolerate the prescription so that they returned after 3 or 4 days.
More presbyopes (88.1%) returned for non-tolerance examinations than pre-presbyopes. These results are comparable to the findings of Constantine-Smith (2002), who found 84% of non-tolerances were for presbyopes. The highest number of non-tolerances in this study was for 50–59 year olds. Patients in the 40 year age group experience loss of accommodation over a period of approximately 20 years (Bennett and Rabbetts, 1989). In this study there were no non-tolerances for patients between 16–19 years; or 90 years and above.
Recent work by Cummings et al. (2007) showed an increased rate of falls in older people, who had their refractive error changed vs a control group. Many of the refractive error changes in that study were >0.75 D, which led the authors to suggest that large prescription changes may increase the risk of falling. As a result, Harley et al., (2007) in their analysis of fall prevention in older people, recommend partially prescribing large refractive error changes to ensure adaptation is as easy as possible. In the present research, there were no non-tolerances to people over the age of 90 years, and only one in the 80–89 age range. This needs to be considered in the context of the relatively small numbers of patients who were prescribed spectacles in these age categories (Figure 2). Additionally, there are several possible reasons why non-tolerances might be less common in older people. Specifically, older patients: might be less sensitive to detecting changes that would have caused a non-tolerance if they were younger; might accept symptoms such as blur as an inevitable effect of ageing; or they might be less likely to return to the practice because of mobility or health problems.
There was an even distribution between men and women returning for non-tolerance examinations, demonstrating that gender does not appear to be a factor in prescription non-tolerance.
In this study, the average rate of prescription non-tolerance, amongst the seven practitioners, was found to be 1.8%. The 95% confidence interval of this estimate is 1.4 to 2.3%. This average rate is slightly less than Riffenburgh et al., (1983) (2.8%) and Mwanza and Kabasele (1998) (2.8%), but slightly higher than the more recent study by Hrynchak (2006) (1.6%). The first two studies by Riffenburg et al. and Mwanza and Kabasele took place in ophthalmological practices. The highest rate from the Mwanza and Kabasele study, which is outside the confidence interval of the present research, may be explained by the small sample size (12 non-tolerance cases out of 432 patients) and the fact that ophthalmologists were refracting. One might expect optometrists, who refract more regularly, to be more accurate than ophthalmologists, for whom refraction is a relatively small part of their work and training. Also, it is possible that patients in an ophthalmological practice had more complex refractive errors and pathologies than those in an optometric practice.
The rates of prescription non-tolerance per practitioner, in this study, ranged from 1.3% to 3.3%. The practitioners were experienced clinicians, averaging 15 years post-qualification. There is no recent comparable study to contrast with our results. Extrapolating from Riffenburgh et al.’s (1983) study (125 non-tolerances from a total of 5467 patients), the three ophthalmologists had 2.6%, 2.2%, and 2.0% of patients returning with complaints with their glasses. This is similar to the results of the present study.
The main reasons for non-tolerance examinations were, in order of decreasing frequency, those related to the prescription (61.0%), those related to the spectacle dispensing (22.0%), pathology related (8.5%), data entry errors (6.8%) and binocular vision problems (1.7%).
Errors relating to the prescription accounted for the majority of non-tolerance examinations (61%). This is similar to Hrynchak’s study where errors, either with incorrect prescription measurement or inability to adapt to the prescription, accounted for 69.3% (Hrynchak, 2006). In this study, 84.4% of prescription enhancements (see below) were within ±0.50 D and 100% within 1 D. This figure agrees with Constantine-Smith (2002) who found, on his practice audit, 87% of recheck (non-tolerance) appointments to be within ±0.50 D. In the control group of 116 patients who collected spectacles in the same period as this study, 79.3% had changes of 0.50 D or less, which is comparable to the non-tolerance group. This finding is not unexpected since Goss and Grosvenor’s (1996) review of the literature on refraction concluded that refraction was reliable to ±0.50 D on 95% of occasions and Miller et al.’s (1997) study showed that errors as small as 0.25 D may not be tolerated. It may therefore be concluded that occasional non-tolerances are to be expected.
Measurements of the spherical part of the refractive error produced the most prescription errors, accounting for 20% of all non-tolerance examinations. Hyperopic patients were more likely not to tolerate a prescription that was too strong (over-plussing). Several studies also agree with these findings (Miller et al., 1997; Atchison et al., 2001; Hrynchak, 2006). Hrynchak found this was the most common error in refractive error measurement. Miller et al.,’s (1997) study also noted that 45% of subjects preferred their ‘correct’ control pair of spectacles to the test pairs with an extra +0.25 and +0.50 D. Atchison et al. (2001) found that the majority of patients (50–80%) preferred lenses without the induced errors in the test spectacles, least tolerated were the +0.50 D monocular and binocular test pairs. In both of these studies, the quality of distance vision was the main factor in determining whether the spectacles were acceptable, however all the subjects were pre-presbyopic and would be expected to be able to accommodate for changes in near and intermediate vision.
Myopic patients, in the present study, were more likely to return for a prescription that was too weak (under-minusing). In general, myopic patients are more likely to notice under-correction, especially during tasks at night, where night myopia occurs (Michaels, 1981). Early presbyopes are the exception to this situation. This group of patients are more likely to notice an over-correction, since this will lead to increasing problems with their near vision (Elliott, 2003).
The second most common reason for a non-tolerance examination was for dispensing errors (22.0%), and as noted above this was probably an under-estimate. Examinations were classified in this category if, after checking the prescription, the remaining problem was felt to be an error with the dispensing of the lens or spectacle frame. This value is higher than Hrynchak (2006), who found 15.3% of non-tolerance examinations were for errors relating to dispensing, either for ‘ineffective education regarding lens design’ or ‘problems with the appliance’. One possible reason for the higher value in this present study is that, during the study period, the dispensing department was under-staffed, with fewer qualified dispensing opticians than normal. This may have resulted in patients being booked for non-tolerance examinations, when ideally the problem should have been resolved, either at the collection of the new spectacles, or by a dispensing optician at a subsequent visit.
Progressive addition lens (PALs) were the lens form that was most often not tolerated, followed by vocational lenses. The literature on dispensing non-tolerances describes more areas for errors in PAL lens dispensing compared with other lens forms (Young, 1984, 1984; McCarthy, 2003; Farrell, 2005). A small proportion of patients have problems adapting to PALs: a survey by Sullivan and Fowler (1989) found that the proportion of unsuccessful wearers was between 11% and 16% (depending on the criterion used), with more problems with the reading portion of the lens than the distance portion. It should be noted that since this research 20 years ago, PAL design has improved so the rate nowadays may be lower than that found by Sullivan and Fowler. However, such a conclusion must be speculative since a PubMed search by the present authors did not identify any more recent research. Single vision lenses only accounted for 2% of non-tolerance cases, despite being 70.5% of the total lenses dispensed over the period. Bifocals accounted for the smallest number of non-tolerance cases, however this lens type only accounted for 6.7% of the total lenses dispensed. Additionally, in the practice, since bifocals are mostly prescribed to existing wearers, one would expect less chance of non-tolerance compared to a lens prescribed more regularly to neophytes, such as PALs. Of the non-tolerance examinations that were for PAL wearers, 75% were either first time PAL wearers or existing PAL wearers using a new PAL lens design, which shows that greater care must be taken with these two groups of patients.
Pathology was the third most common cause of non-tolerance accounting for 8.5% of non-tolerance examinations. This is similar to the proportion found in Hrynchak’s (2006) study (9.3%), but lower than Riffenburgh et al.,’s (1983) study, where 16.8%, or 21 of the 125 return examinations were for cataract. Riffenburgh et al.’s study, however, took place at a non-surgical ophthalmological practice, with older patients, and where more complex refractive errors and pathologies might be expected than in an optometric practice.
It is well known, in optometric practice, that patients with cataract tend to be more difficult to refract and show greater variability (Legge et al., 1987; Leinonen et al., 2006). Greater care needs to be taken before deciding whether to issue a change in prescription. Leinonen et al. (2006) suggest that, compared with a patient with normal vision, it would be advisable to double the amount before a change in prescription is issued.
The other ‘pathological’ cause of non-tolerance was moderate dry eye. This patient experienced intermittent blurring and the non-tolerance was dealt with by prescribing ocular lubricants, with no change to the prescription. Montes-Mico et al. (2004) also observed that dry eye patients experience increased aberrations, or changes which reduce the quality of the retinal image, compared to normal patients.
Data entry errors only occurred four times during the study period. Other studies have also highlighted this cause of non-tolerance (Veasey, 1946; Priest, 1979; Hrynchak, 2006; Steele et al., 2006). In the present study, all prescriptions were entered into the examinations module of the practice computer system, and then dispensed from the dispensing module of the same system; no paper records were used. Incorrect entry was the cause twice in the prescription software module and twice in the spectacle dispensing module. Steele et al. (2006) classified all errors that occur in optometric practice and found that errors in transcription were the most common of the optical prescription errors.
Binocular vision problems accounted for one non-tolerance case only. This was a case of convergence insufficiency associated with aging. The management of the non-tolerance involved prescribing a prismatic correction and exercises were recommended. This convergence insufficiency non-tolerance has also been noted in another non-tolerance study (Veasey, 1946). Binocular vision accounted for only 1.7% of non-tolerance examinations; this is much lower than the Hrynchak (2006) study (10.3%). The Hrynchak study, however, occurred in a region where patients had to wait 6 months between cataract surgeries, which resulted in several non-tolerances due to induced anisometropia following the unilateral cataract surgery.
Inevitably, there are some limitations to the present work. As noted at the beginning of the Discussion, patients with tolerance issues were first seen by the dispensing team and the research is restricted to cases that the dispensing team had not been able to resolve. Although in keeping with most previous research, this means that the present research will underestimate the true prevalence of dispensing errors and data entry errors. A second limitation is that it is assumed in this research that any prescriptions re-issued at the non-tolerance examination were the correct final prescription. Although patients were not contacted to check that they were happy with the final prescription, no patients returned a third time despite it being made clear to all patients that they should return if there were any further problems and that they would receive a full refund if they were still unhappy with their spectacles.
Another limitation of any research assessing optical prescriptions is that many clinicians sometimes make adjustments to the refractive error that they find in their subjective refraction before arriving at a final prescription that is issued to the patient. The types of adjustments that are made and rationale behind them have been recently reviewed by Elliott (2008). Enquiries were made of the optometrists who participated in the present research, but this revealed that although all the optometrists do commonly make minor adjustments to their subjective findings before prescribing, there are no fixed ‘rules’ that could be explicitly specified. As noted by Elliott (2008), there are several influences that are likely to act on these clinical judgements and the effect of such prescribing adjustments on the prevalence of non-tolerance would be an interesting topic for future research.
From this study there are many points that are of direct relevance to optometric practice. The main clinical implications are listed in Table 4. By highlighting the main types of prescription errors and ways to avoid them, practitioners can learn how to minimise them in future.
Table 4. Clinical implications from this study
|Do not over-plus hyperopes||These cases will be aware of poor quality distance vision|
|Do not under-minus myopes||Problems with night vision|
|Do not overestimate the near/intermediate addition||Patients are less tolerant of over-correction|
|Investigate carefully each patient’s working distance for reading or computer use||It is difficult for patients to assess this correctly in the testing room environment, so a pre-exam questionnaire might be useful|
|If you use a refractor head (phoropter), make an allowance for BVD||Refractor heads can lead to over-correction in myopes and under-correction in hyperopes|
|Take care with testing distance, especially with 3–4 m projection charts||A 6 m chart overcorrects by +0.167 D and a 4 m chart overcorrects by +0.25 D (so need to reduce prescription by −0.25 D) |
|Pay particular attention to changes in prescription with presbyopes||More non-tolerances likely, especially 50–59 age group|
|Take care when prescribing to new multifocal wearers or when changing the design of existing multifocal wearer||These two groups of multifocal wearers are more sensitive to changes and more prone to non-tolerance|
|Take greater care when prescribing to patients with cataract||Patients are more difficult to refract and show greater variability. Compared with a non-cataract patient, consider doubling the amount before a change in prescription is issued|
|Make sure records are completed fully and accurately||Many non-tolerances are simply the result of incorrect data entry|
Throughout this and other studies the term ‘non-tolerance’ examination is used. This term implies that the problem is with the patient, not the practitioner, which it is suggested is pejorative. In this study, the category most accurately describing patient non-tolerance is adaptation problems and to a lesser degree pathology. Other categories, such as errors with sphere, errors with cylinder, near/intermediate addition and data entry error, relate to practitioner errors, albeit small errors, the majority 0.50 D or less. In future it may, therefore, be more appropriate to replace ‘non-tolerance’ with a more neutral term such as ‘refractive correction adverse event’. Alternatively, the literature on laser refractive surgery commonly uses the more positive description of ‘enhancement’ (Benito-Llopis et al., 2009).
In summary, non-tolerance examinations comprised a small percentage (2.0%) of eye examinations during the study period and the average rate of prescription non-tolerance, amongst the seven practitioners was found to be 1.8%. In a climate where clinical governance and auditing are increasingly important, an understanding of the norms for prescription non-tolerance can help practitioners to determine best practice.
The majority of patients can be helped by either a small change to their prescription or their spectacles, in most cases 0.50 D or less. Sometimes, as is the case with adaptation problems and pre-diagnosed pathology, an explanation of the cause is enough to resolve the problem.