Error classification in community optometric practice – a pilot project

Abstract

Objective:  A pilot study was conducted to undertake a baseline assessment of errors reported in community optometric practice. The feasibility and acceptability of a method for recording staff-reported errors in optometric community practice was investigated.

Design:  An anonymous self-reporting system was introduced in order to collect information regarding errors/untoward events in community optometric practice.

Setting:  UK community optometric practice.

Main outcome measures:  Classification of errors according to a previously published study pertaining to general medical practice in the same geographical area.

Results:  Thirty-six notebooks were distributed to 10 participating community optometric practices. At the end of the 1 month study period the note books were returned and the 439 entries made were classified into seven categories: optical prescriptions (18.2%), communication (35.5%), administrative (15%), appointments (2.3%), equipment (11.9%), clinical (10.3%) and other (6.9%).

Conclusion:  A previously developed classification of errors in general medical practice was found to be equally applicable to community optometric practice. This study forms the basis of providing an acceptable and practical methodology, which can be applied by Primary Care Trusts (PCTs) when developing their risk management strategy to include optometry.

Introduction

Risk management is an essential element of clinical governance (CG) (Department of Health, 1999) and the importance of detecting and recording errors as a key step in learning from experience has been emphasised (Department of Health Expert Group, 2000). Modern health care is characterised by reliance on human operators who increasingly work with complex technology (Allnut, 1987; Meurier, 2000). Although not specifically mentioned, these previous studies apply equally to community optometric practice, which plays a key role in primary eye care. Arguably errors are inevitable, but in certain circumstances may have serious consequences (Barach and Small, 2000; Elder and Dovey, 2002).

Industries such as aviation have for decades had well developed methods for documenting and investigating risk that allow systematic efforts to reduce the frequency and severity of adverse events (Helmreich, 2000). In recent years risk management in health care has improved enormously but risk management in community optometric practice has remained much less well developed (Elder and Dovey, 2002). Good risk management practices have important financial implications for health care providers, including optometrists. In 1998/1999, medical litigation cost the NHS £400 million, with an estimated further £2.4 billion in potential liability (Department of Health Expert Group, 2000). In optometric practice there has also been a steep rise in litigation costs in recent years (Warbuton, 2004).

Clinical governance has become well established in secondary care. Those responsible for CG in primary care have focussed until recently on incidents with the greatest potential for leading directly to an adverse event, even though these are rare (Sanders and Esmail, 2003). Much of the emphasis has therefore been placed on general medical practice (Dovey et al., 2002; Makeham et al., 2002; Tamuz et al., 2004). Community optometric practice has largely been ignored, despite its important role in primary eye care (College of Optometrists, 2002). In England, Primary Care Trusts (PCTs) are National Health Service organisations each serving a population of approximately 100 000, and responsible for improving health care in local areas. PCTs now control the budgets from which NHS eye examination fees are paid to community practice optometrists. In 2003/2004 the number of NHS eye tests paid for by Sunderland Teaching PCT was 55 510 and nationally the total was over 11.5 million (Department of Health, 2004). PCTs have a responsibility for securing the provision of a wide range of services including commissioning of acute and specialist services as well as family health services in primary care. As a means of ensuring high standards in community optometric practice, PCTs will need to develop ways that can be used to maintain and develop standards applicable to this discipline.

Studies in primary medical care have described widely varying rates of clinical and organisational error and provided systems for their classification. None of these studies have included community optometric practice.

The aim of this pilot study was to undertake a baseline assessment of errors reported in community optometric practice. The feasibility and acceptability of a method for recording staff-reported errors in community optometric practice were investigated. To the authors’ knowledge no such similar research has been conducted into community optometric practice previously, and this gives a snap shot of the current position.

Methods

Using a previously described method of self-reporting (Rubin et al., 2003), a pilot study was conducted during April 2004 involving 10 community optometric practices in Sunderland and Durham. Prior to agreeing to be involved in the study all staff at each practice had been briefed and the nature of the study explained. Of the 10 practices involved a total of six were independent practices owned by individuals or partners and four were multiples, i.e. one of a number of optical practices belonging to a large company. The support of the local optometric committees was obtained for the study.

Participants were given no advice prior to the study on what constituted an error and this was left to individuals’ discretion. All staff (i.e. professional, administrative and support) in each practice were encouraged to take part and they were asked to record free text descriptions of events in blank notebooks for a period of 4 weeks. They were asked to record all errors no matter how seemingly trivial, as they occurred and at every occurrence. Notebooks were placed in all clinical as well as administrative areas rather than giving notebooks to individuals, in order to promote anonymity. Every effort possible was made to assure participants that all information recorded was confidential and non-attributable. A total of 36 notebooks were distributed, the number for each practice varied according to its size.

At the end of the 4 week study period the notebooks were returned to two of the authors (CFS and SF) who analysed the recorded events using a previously described taxonomy as follows:

• 1
  Prescriptions – optical and drug related
• 2
  Communication
• 3
  Administrative
• 4
  Equipment
• 5
  Clinical
• 6
  Appointments
• 7
  Other

Results

A total of 15 practices were actually approached but five declined to take part. Reasons given were that they were ‘too busy’, ‘not really interested’ and/or concerns were raised that information divulged might not be kept confidential and could later possibly be used to their detriment. Participating optometric practices had a median patient base of 6543. During the study period a total of 115 people worked in the 10 practices: 24 registered optometrists (of a total of 71 in Sunderland and 74 in Durham), 3 pre-registration optometrists, 8 dispensing opticians and 80 optical assistants/administrative staff.

Of the 36 notebooks distributed, 36 (100%) were returned to the authors at the end of the study period and a total of 439 entries had been recorded (Table 1). The median number of entries per practice was 31; range: 6–118. The larger multiple practices had relatively the greatest number of entries.

Table 1.  Summary of entries recorded compared with findings from Rubin et al. (2003)

Classification	Entries (n = 439)	Percentage (%)	Rubin et al. (%)
Optical prescriptions	80
Many transposition errors		18.2	Prescriptions 42.2%
Illegible forms
Various details not recorded e.g. bifocal measurements
Non-tolerances
Patient worn wrong spectacles for 4 years – unaware!
Communication	156
63 referrals only 14 replies		35.5	30.0
Missing records (n = 85)
Illegible writing (in e.g. clinic records)
Too many product promotions running at once
Failure to explain clinical investigations e.g. visual fields – errors caused
Patients not kept informed of delays
Pt report use of Viagra optometrist thought they were joking
Administrative	66
Locum optometrists not completing forms correctly		15	21.7
Patient demographic data errors
System errors in processing spectacle and contact lens orders
Equipment	52
Injuries sustained whilst undertaking spectacle frame repairs and adjustments, e.g. screwdrivers gouging fingers		11.9	16.3
Computer problems (hardware and software)
Technical equipment failures, e.g. auto-refractor and auto focimeter malfunctions
Fell off ladder in stock room
Chair cabinet related injuries
Clinical	45
Relatively minor admissions		10.3	2.6
Wrong contact lens solutions supplied (peroxide instead of multipurpose solution
Clinical tests forgotten, e.g. tonometry, visual fields and amsler
Records not completed
Disagreement on grading findings which affect referral criteria
Pre-reg supervisor off sick!
Appointments	10
Appointments given on the wrong day		2.3	6.7
Patients presenting on the wrong day
Other	30
Dealing with abusive patients		6.9	2.2
Patients tripping on door steps
Patients having difficulty climbing in-practice staircases
Various items of equipment dropped into patients’ private places
Optometrist's inner thighs being fondled during ophthalmoscopy!
Total	439	100	100

Communication errors were the most commonly reported and accounted for 156/439 (35.5%) of the total. Of the communication errors, the largest number recorded was for missing case notes [85/156 (19.7%)]. Optical prescription errors accounted for 80/439 (18.2%) with errors in transposition and illegible forms being the most common, followed by errors in measuring spectacle dispensing details, e.g. bifocal segment heights and inter-pupillary distances. Administrative errors accounted for 66/439 (15%) with most of these relating to errors in processing optical orders or certain staff not completing forms correctly and/or completely. The most common equipment errors 52/439 (11.9%) were computer related or due to technical difficulties with specialist technical equipment such as auto refractors, auto lensmeters (focimeters) and visual field instrumentation. Only 45/439 (10.3%) were clinical errors with inaccurate note keeping and discrepancies in grading clinical findings being the main source reported. As in the study by Rubin et al. (2003) a further category called ‘other’ was included which was used to record any other error that did not fit into the other categories. The entries included in this category were wide ranging (see Table 1).

Discussion

This is the first study of its kind to address risk management issues in community optometric practice. The error classification system developed by Rubin et al. (2003) in general medical practice, was equally applicable to community optometric practice. The proportion of errors by this classification system in community optometric practice was also similar to general medical practice. The methodology used in this study was based on that described by Rubin et al. (2003). This was just as feasibly applied to optometric practice as it had been to general medical practice.

Other large studies have produced more complex classification systems with as many as 90 categories (Runciman et al., 1998) in contrast to the seven categories used here. In primary care a classification system with 90 categories would be too unwieldy. Conversely another recent study classified errors into just two broad categories: ‘process errors’ (including administration, investigations, treatments, communication and payments) and ‘knowledge and skills’ errors (Tamuz et al., 2004).

No definition of an error was given prior to this study. Other studies however have used a broad definition of an error as ‘an event that was not completed as intended and/or meant that work was disrupted in some way’; which was based around a definition proposed by Reason (2000). By deliberately not giving a prior definition of an ‘error’, this gave a valuable insight into what individuals actually perceived as errors and also helped reduce the need for participants to analyse events too closely before recording them. This made the recording process quick and relatively uncomplicated and not too disruptive. The authors believe that using several notebooks in each practice in order to encourage anonymity helped reduce the fear of reprimand and therefore lead to the high number of entries recorded (n = 439) during the study period.

Previous studies have demonstrated that under-reporting of events is a problem no matter what methods have been employed and may range in extent from 50 to 96% (Barach and Small, 2000). Despite our best efforts in this study it is almost certain that the number of errors was also under-reported. This is particularly so for clinical errors where optometrists’ fears of disciplinary action or litigation become more apparent (Weingart et al., 2001). Conversely in some cases more than one individual may have reported the same incident (Bhatia et al., 2003).

This was a pilot study limited to the Sunderland and Durham area of the North East of England. There is no reason to believe that the participating optometric practices were not representative of the type of practices in the area or in other parts of the country. The multiples were indeed the larger practices with significantly more optometrists and therefore more support staff compared with the independents. Consequently the larger practices may have affected the overall results. The independent practices are probably organised differently in many respects compared with the multiple practices, although this is not to suggest that the quality of patient care is in any way inferior. This is similar to single GP practices compared with large GP practices with several partners (Hippisley-Cox et al., 2001). Certainly independent practices are just that; there is more flexibility to exercise clinical and professional freedom compared with a multiple, where there is an expectation for individuals to abide by corporate policies and procedures. Most of the practices that did not wish to participate were smaller independent practices. With medical practices (Rubin et al., 2003), this was mainly explained by concerns around additional work load but also perceived concerns around anonymity in reporting errors that could only be attributed to one individual rather than several.

In this study the most commonly reported errors were in communications which is in contrast to the study by Rubin et al. (2003) who found prescription errors to be the largest category. Administrative and prescription errors were also common, the later being mainly due to mistakes in transcription. This indicates the importance of final checking of prescriptions (Chua et al., 2003). That these three categories were the most common may reflect the fact that the largest group of staff in this study were non-clinical. Certainly these types of errors may also be thought to be less hazardous and therefore less threatening to report. Due to the anonymous nature of this study it was not possible to compare reporting rates between clinical and non-clinical staff. Further information on the outcomes of errors was not specifically sought either which makes any significant event audit or root cause analysis to investigate potential solutions very limited.

Communication errors accounted for 156/439 (35.5%) of the total. One of the most significant concerns in this area was the high rate of optometric referrals made to secondary care that did not generate any feedback to the optometrist from the ophthalmologist. Therefore the optometrist has no way of knowing whether the referral was appropriate or not and what the outcome of treatment (if any) was. This is an indictment on current relationships between optometric primary care and secondary ophthalmologyical care. Without feedback optometrists are unable to refine their referrals and learn from these outcomes and therefore improve the quality and appropriateness of referrals made. This is one specific area that requires urgent attention by all concerned in order to promote seamless care for the patient.

Modern optometric practice increasingly relies on the use of a range of sophisticated and technologically advanced ophthalmic instrumentation. The instruments are however not without their problems and significant faults were reported with certain items of equipment such as auto-refractors and auto-lensmeters. All the participating practices were computerised and problems encountered involved both hardware and software.

The study period was for 1 month, which is a relatively short period of time compared with other larger studies in this field. However a month was considered to be a reasonable length of time to achieve the aims whilst maintaining support and enthusiasm for the study by those involved. In contrast, Rubin et al. (2003) used a 2 week period in their study to maintain enthusiasm by those concerned.

Finally, acceptability of this study was not formally tested. However towards the end of the study period each participating practice was contacted by telephone to remind them to return the notebooks by the due date. When asked how each practice had found filling in the notebooks, anecdotally at least, no negative comments were received from any practice.

Conclusions

Our findings give a snapshot of the situation regarding error reporting and their classification in primary optometric care during the study period. In many aspects they mirror the situation found in general medical practice. The anonymous self-reporting methodology used was acceptable and useful in identifying specific problems in optometric practice that require action. By highlighting common areas of concern, this will help PCTs to decide how to develop and implement risk management strategies in the context of a non-punitive culture (Department of Health, 2001).