770
Journal of the Royal Society of Medicine Volume 83 December 1990
General practitioner referrals to a-neye hospital: a standard referral form
N P Jones FRcs I C Lloyd FRCS J Kwartz DO University Department of Ophthalmology, Manchester, Royal Eye Hospital Oxford Road; Maihester Mi1 9WH Kewords: general practitioner referral; hospital eye service; standard-ireferrl ferm
Summary To assess general practitioner (GP) referrals to an eye hospital, 500 consecutive referral letters were analysed for content, diagnosis, ocular examination, and medical and drug history. Inadequate infortion was provided. Visual acuity was measured by general practitioners in only 3.7% of referrals. The standard of ocular examination compared unfavourably with that of ophthalmic opticians. A standard ophthalmic referral form is proposed, and referral protocols are desirable. Undergraduate education in ophthalmology is inadequate and requires more curricular time. Introduction The quality of communication between the hospital services and general practitioners has recently come under scrutiny1-10. The discharge letter from a hospital unit to a general practitioner is an example of a method of communication often considered to be in need of improvement8-10. Although the principles of referral from general practitioner (GP) to hospital specialist have been addressed2-5, the quality of the referral letter itself is not so frequently considered6. Ophthalmology is a busy outpatient specialty with, in some cases, long waiting times for a first hospital appointment. With increasing demand, and with a high prevalence of ocular disease in an ageing population"1, such problems are unlikely to decrease. Appropriate, concise and informed referrals from GP to hospital will do much to- smooth the consultation process. At Manchester Royal Eye Hospital, incoming referral letters are received centrally in the medical records department and distributed to consultant firms. The letters are scrutinized by the consultant or senior registrar, and a decision on priority is made for the patient's first appointment, based upon the clinical information contained in the letter. The quality of referral letters are extremely variable. This study quantifies this variability and comments on the effects upon patient triage. Protocols for referral information are suggested, and the introduction of a standard referral letter is proposed.
Methods Five hundred consecutive patient referrals made by post to the Manchester Royal Eye Hospital were analysed. Patients referred urgently with letters by hand to the Accident & Emergency Department were excluded, as were internal referrals from other hospital doctors. Referrals to the hospital were initiated either by the GP or by an ophthalmic optician (00) via referral form GOS 18. The content and quality ofthe referral
letters were analysed in terms of presentation, information provided and the need for priority. Comparison was made between the information presented by GP and 00 where appropriate. No comparison was made between the su s diagnis of the referring doctor or optician, and the final diagnosis of the ophthalmologist. The ophthalmologist scrutinizing the referrals does not have this luxury and must base his decisions on the information provided by the referring source. Form GOS 18 refers a patient from 00 to GP. The form contains a section for completion by the GP if referral to an ophthalmologist is considered necessary. The provision of further relevant information by the GP in this circumstance, was analysed. Results All 500 patients were referred to the hospital by a GP. Of these referrals, 245 (49.0%) were initiated by the GP and 255 (51.0%) by an ophthalmic optician. A diagnosis or suggested diagnosis was made in 409 cases (81.8%). Table 1 shows the spectrum of diagnoses Tablek . Diagnoses requiring examination ofeyelids, adnexae, anterior segment or posterior segment Diagnosis
Number Number (GP) (00)
Diagnoses requiringqexandrton of eyelids and adne Eyelid lesions for minor surgery Limal problems Problems with ocular prostheses
79 12
0
3
2 0
of anterior segment: Coijunctival and corneal problems 20 38 Cataract
14 74
.
Diagnoses requiring examination
Diagnoses requiring examination of posterior segment: Possible glaucoma Macular degeneration Diabetic retinopathy Hypertensive retinopathy Other fundus problems
2 2 2 0 0
47 18 6 5 14
23 1 3 50 10
20 2 3 41 9
Diagnoses requiring neurological or orthoptic aesment Orthoptic problems Ptosis Intracranial problems No diagnosis made
Miscellaneous
0141-0768/90/
120770.03/02.00/0 0 1990 The Royal Society of Medicine
Journal of the Royal Society of Medicine Volume 83 December 1990 Table 2. Diagnosis in referral letter: patients with reduced vision
Diagnosis
Number (GP) Number (00)
2 Amblyopia 35 (44.3%) Cataract 2 Corneal problems 2 Diabetic retinopathy 2 Glaucoma 2 Macular degeneration Retinal vascular problems 1 2 Others No diagnosis made 31 (39.2%)
Total
79
3 70 (54.2%) 4 3 2 15 3 6 23 (17.8%) 129
Table 3. Ocular examination performed by general practitioners (GP) and ophthalmic opticians (00) as recorded in the referral letter
Form of examination
Measurement of visual acuity External/anterior segment examination Examination of fundus Comprehensive examination including acuity, anterior and posterior segment No record of examination
Patient
Patient
nos. (GP)
nos. (00)
9 (3.7%)
183 (71.7%)
131 (53.5%) 129 (50.6%)
15 (6.1%) 3 (1.2%)
74 (29.0%) 27 (10.6%)
107 (43.7%)
42 (16.5%)
for referrals initiated by GP and 00. For those patients with symptoms of visual deterioration, the suggested diagnoses are shown in Table 2. Where definite evidence was included in the referral letter of some form of ocular examination having taken place, this information was recorded. The examinations performed by GP and 00 are recorded in Table 3. General practitioners measured the visual acuity in only 3.7% of referrals initiated by themselves, and for patients with visual loss, the proportion rose to 6.3%. The presentation ofreferral from the GP varied. In 313 cases (62.6%) there was a typed letter. In 92 cases (18.4%) the referral was in the form of a legible handwritten letter. It was not possible to read the letter in 22 cases (4.4%) owing to illegibility or incomprehensibility. In 305 referral letters (61.0%) the patient's symptoms were recorded or were implied by the diagnosis. Information on the duration of symptoms was provided by the GP in 54 cases (10.8%). A past ocular history was provided in 50 cases (10.0%) and information on topical treatment already prescribed in 30 (6.0%). Some information on general health was included in 112 referrals (22.4%) and this proportion rose to 27.5% for those 127 patients complaining of loss of vision. A note of currently prescribed drugs was included in 67 cases (13.4%). For those 255 patients referred initially by the 00, an accompanying letter was sent by the GP in 74 (29.0%) cases in addition to the form GOS 18. In 132 referrals (51.8%) form GOS 18 alone was sent, with or without further information included. In 49 (19.2%)
instances a letter was sent but form GOS 18 was not enclosed. For those patients referred initially by the 00, the following additional information was supplied by the GP: ocular history 24 (9.4%); present ocular medication 2 (0.8%); past medical history 62 (24.3%); drug history 36 (14.1%). In 166 referrals (65.0%), no clinical patient information was supplied by the GP.
Discussion The spectrum ofdiagnoses shown in Table 1 illustrates patients' perceptions of the roles of GPs and opticians. It also reflects opticians' greater knowledge of ophthalmology and greater skill in ophthalmoscopy. Both professionals have an active part to play in the primary care of ocular problems. Ocular complaints account for 2.7% of all GP consultations12, yet undergraduate medical education in ophthalmology does not adequately reflect this. There is much room for improvement. Clearly, a referring doctor is selective about what is included in a referral letter. For this reason it should not be assumed that failure to include evidence of an ocular examination, indicates that no examination has taken place. However, for the ophthalmologist such information is important and the inclusion of normal findings is relevant. It is therefore disappointing that 43.7% of referrals from GPs contained no evidence of an ocular examination, and that only 6.1% of patients from this source had their optic discs or maculae examined. It is also of some concern that only 29% of patients referred by QOs included any comment on the posterior segment, despite the fact that examination of the optic discs is a statutory part of a refraction test. The measurement of visual acuity is, for the ophthalmologist and optician, the first part of any ophthalmic examination. This need not be the case for the general practitioner, but should be mandatory if the patient complains of visual loss or other visual symptoms. The test is not time-consuming and every general practice should be able to provide a Snellen chart for this purpose. It is of concern that even for patients with visual loss, only 6.3% of patients had visual acuity measured by their GP. The figure of 71.7% for opticians was surprisingly low. The importance of providing a medical background for a patient with visual loss was appreciated in only the minority ofreferrals (27.5%). In only 13.4% was any drug information provided. Information on currently prescribed drugs is of great importance in any referral, and no less so because the recipient is an ophthalmologist. Many ophthalmological patients are elderly, and it is in this age group where much inaccuracy in drug reporting is found7. Some drugs can cause ocular disease (eg prednisolone, chloroquine). Others can interact with topical ophthalmological drugs (eg Verapamil with ,8-blockers) whereas some can potentiate drugs used in ophthalmological practice (eg potassium-losing diuretics and aceta7nlamide). Some have important implications for surgery, and this includes minor eyelid surgery (eg warfarin). Drug information includes topical treatment, and in only 6% of this study was such information included. We are aware that by no means all patients will have used topical treatment, but it is common for such information to be absent when treatment has been given. The information is important especially where an antibiotic has already
771
772
Journal of the Royal Society of Medicine Volume 83 December 1990
been tried, and is crucial if a topical steroid has been prescribed. For those patients initially referred by their optician, 65% had no further information supplied before the referral was forwarded to the ophthalmologist. It was also disappointing that in 19.2%, form GOS 18 was not sent on by the GP. With this referral route, both optician and GP have impQrtant information to divulge. Form GOS 18 contains a refraction and details of an ophthalmic examination which will be comprehensible to an ophthalmologist but often not to a GP. This form should always be included. Similarly, many patients with ocular complaints will have underlying medical problems unknown to their optician, which can be placed into perspective by their general practitioner. This opportunity should not be missed. Suggestions that opticians refer directly to ophthalmologists should be resisted for this reason, except in the occasional purely ophthalmic emergency such as acute glaucoma. The question of screening for ocular diseases, especially glaucoma, has been raised in the past and the suggestion has been made that general practitioners have a role to play13. Opticians are however better placed to do this. They are better trained in ophthalmology (though this is a condemnation of undergraduate medical training), are more likely to be accurate in their diagnosis614"5 and have access to tonometry and visual field testing, which is necessary to make a better assessment of the patient than is possible by mere examination of the optic discs16. That opticians are already more active in this field is supported by this study, in which GPs referred two glaucoma suspects, and opticians 47. In an attempt to improve the quality and layout of information provided in a hospital discharge Practice stamp
Date
Dear
DOB
Re:
Address:
Occupation: Now Patient
Eye Hosp No: Symptoms:
Duration: Treatment:
Exasination:
Best Visual acuity
R
L
Diagnosis:
Hedical History:
Current drugs:
Priority:
Routine
Some priority
Urgent
Reason:
Yours
Figure 1. Proposed format for a standard referral letter
summary, a standard format has been suggested8. The concept can equally be applied to a referral letter, and an example is given in Figure 1. The letter contains practice and patient details, and is followed by a concise list of information which would enable rapid and accurate assessment of the patient's needs by the ophthalmologist. These details include some statement on symptoms and their duration, the essentials of ocular examination, and supporting information on past medical history and prescribed drugs. The widespread usage of such forms, together with the mutual development of referral protocols as suggested by Marinker et aL3 can only improve the standard ofreferral and therefore the service provided to patients. The provision of a similar section along these lines would also enhance the GOS 18 referral form. Acknowledgments: We are grateful for the willing help of the staff of the Medical Records Department, Manchester Royal Eye Hospital. References 1 Fair JF. Hospital discharge and death communications. Br J Hosp Med 1989;42:59-61 2 McGlade KJ, Bradley T, Murphy GJ, Lundy GP. Referrals to hospital by general practitioners: a study of compliance and communication. BMJ 1988;297: 1246-8 3 Marinker M, Wilkin D, Metcalfe DH. Referral to hospital: can we do better? BMJ 1988;297:461-4 4 Hull FM, Westerman RF. Referral to medical outpatients department at teaching hospitals in Birmingham and Amsterdam. BMJ 1986;293:311-14 5 Coulter A, Noone A, Goldacre M. General practitioners' referrals to specialist outpatient clinics. BMJ 1989:304-8 6 Harrison RJ, Wild JM, Hobley AJ. Referral patterns to an ophthalmic outpatient clinic by general practitioners and ophthalmic opticians and the role of these professionals in screening for ocular disease. BMJ 1988; 297:1162-7 7 Gilchrist WJ, Lee YC, Tam HC, MacDonald JB, Williams BO. Prospective study of drug reporting by general practitioners for an elderly population referred to a geriatric service. BMJ 1987;294:289-91 8 Penney TM. How to do it. Dictate a discharge sutmmary. BMJ 1989;28:1084-5 9 Penney TM. Delayed communication between hospitals and general practitioners: where does the problem lie? BMJ 1988;297:28-9 10 Mageean RJ. Study of "discharge communications" from hospital. BMJ 1986;293:1283-4 11 Gibson JM, Rosenthal AR, Lavery J. A study of'the prevalence of eye disease in the elderly in an English community. Trans Ophthalmol Soc UK 1985;104: 196-203 12 Dart JK. Eye disease at a community health centre. BMJ 1986;293:1477-80 13 Hitchings RA. Visual disability and the elderly. BMJ 1989;298:1126-7 14 Clearkin L, Harcourt B. Referral pattern of true and suspected glaucoma to an adult ophthalmic outpatient clinic. Trans Ophthalmol Soc UK 1983;103:284-7 15 Brittain GP, Austin DJ, Kelly SP. A prospective survey to determine sources and diagnostic accuracy of glaucoma referrals. Health Trendls 1988;20:43-4 16 Wood CM, Bosanquet RO. Limitations of direct ophthalmoscopy in screening for glaucoma. BMJ 1987;294: 1587-8
(Accepted 8 May 1990)
Journal of the Royal Society of Medicine Volume 83 December 1990
General practitioner referrals to a-neye hospital: a standard referral form
N P Jones FRcs I C Lloyd FRCS J Kwartz DO University Department of Ophthalmology, Manchester, Royal Eye Hospital Oxford Road; Maihester Mi1 9WH Kewords: general practitioner referral; hospital eye service; standard-ireferrl ferm
Summary To assess general practitioner (GP) referrals to an eye hospital, 500 consecutive referral letters were analysed for content, diagnosis, ocular examination, and medical and drug history. Inadequate infortion was provided. Visual acuity was measured by general practitioners in only 3.7% of referrals. The standard of ocular examination compared unfavourably with that of ophthalmic opticians. A standard ophthalmic referral form is proposed, and referral protocols are desirable. Undergraduate education in ophthalmology is inadequate and requires more curricular time. Introduction The quality of communication between the hospital services and general practitioners has recently come under scrutiny1-10. The discharge letter from a hospital unit to a general practitioner is an example of a method of communication often considered to be in need of improvement8-10. Although the principles of referral from general practitioner (GP) to hospital specialist have been addressed2-5, the quality of the referral letter itself is not so frequently considered6. Ophthalmology is a busy outpatient specialty with, in some cases, long waiting times for a first hospital appointment. With increasing demand, and with a high prevalence of ocular disease in an ageing population"1, such problems are unlikely to decrease. Appropriate, concise and informed referrals from GP to hospital will do much to- smooth the consultation process. At Manchester Royal Eye Hospital, incoming referral letters are received centrally in the medical records department and distributed to consultant firms. The letters are scrutinized by the consultant or senior registrar, and a decision on priority is made for the patient's first appointment, based upon the clinical information contained in the letter. The quality of referral letters are extremely variable. This study quantifies this variability and comments on the effects upon patient triage. Protocols for referral information are suggested, and the introduction of a standard referral letter is proposed.
Methods Five hundred consecutive patient referrals made by post to the Manchester Royal Eye Hospital were analysed. Patients referred urgently with letters by hand to the Accident & Emergency Department were excluded, as were internal referrals from other hospital doctors. Referrals to the hospital were initiated either by the GP or by an ophthalmic optician (00) via referral form GOS 18. The content and quality ofthe referral
letters were analysed in terms of presentation, information provided and the need for priority. Comparison was made between the information presented by GP and 00 where appropriate. No comparison was made between the su s diagnis of the referring doctor or optician, and the final diagnosis of the ophthalmologist. The ophthalmologist scrutinizing the referrals does not have this luxury and must base his decisions on the information provided by the referring source. Form GOS 18 refers a patient from 00 to GP. The form contains a section for completion by the GP if referral to an ophthalmologist is considered necessary. The provision of further relevant information by the GP in this circumstance, was analysed. Results All 500 patients were referred to the hospital by a GP. Of these referrals, 245 (49.0%) were initiated by the GP and 255 (51.0%) by an ophthalmic optician. A diagnosis or suggested diagnosis was made in 409 cases (81.8%). Table 1 shows the spectrum of diagnoses Tablek . Diagnoses requiring examination ofeyelids, adnexae, anterior segment or posterior segment Diagnosis
Number Number (GP) (00)
Diagnoses requiringqexandrton of eyelids and adne Eyelid lesions for minor surgery Limal problems Problems with ocular prostheses
79 12
0
3
2 0
of anterior segment: Coijunctival and corneal problems 20 38 Cataract
14 74
.
Diagnoses requiring examination
Diagnoses requiring examination of posterior segment: Possible glaucoma Macular degeneration Diabetic retinopathy Hypertensive retinopathy Other fundus problems
2 2 2 0 0
47 18 6 5 14
23 1 3 50 10
20 2 3 41 9
Diagnoses requiring neurological or orthoptic aesment Orthoptic problems Ptosis Intracranial problems No diagnosis made
Miscellaneous
0141-0768/90/
120770.03/02.00/0 0 1990 The Royal Society of Medicine
Journal of the Royal Society of Medicine Volume 83 December 1990 Table 2. Diagnosis in referral letter: patients with reduced vision
Diagnosis
Number (GP) Number (00)
2 Amblyopia 35 (44.3%) Cataract 2 Corneal problems 2 Diabetic retinopathy 2 Glaucoma 2 Macular degeneration Retinal vascular problems 1 2 Others No diagnosis made 31 (39.2%)
Total
79
3 70 (54.2%) 4 3 2 15 3 6 23 (17.8%) 129
Table 3. Ocular examination performed by general practitioners (GP) and ophthalmic opticians (00) as recorded in the referral letter
Form of examination
Measurement of visual acuity External/anterior segment examination Examination of fundus Comprehensive examination including acuity, anterior and posterior segment No record of examination
Patient
Patient
nos. (GP)
nos. (00)
9 (3.7%)
183 (71.7%)
131 (53.5%) 129 (50.6%)
15 (6.1%) 3 (1.2%)
74 (29.0%) 27 (10.6%)
107 (43.7%)
42 (16.5%)
for referrals initiated by GP and 00. For those patients with symptoms of visual deterioration, the suggested diagnoses are shown in Table 2. Where definite evidence was included in the referral letter of some form of ocular examination having taken place, this information was recorded. The examinations performed by GP and 00 are recorded in Table 3. General practitioners measured the visual acuity in only 3.7% of referrals initiated by themselves, and for patients with visual loss, the proportion rose to 6.3%. The presentation ofreferral from the GP varied. In 313 cases (62.6%) there was a typed letter. In 92 cases (18.4%) the referral was in the form of a legible handwritten letter. It was not possible to read the letter in 22 cases (4.4%) owing to illegibility or incomprehensibility. In 305 referral letters (61.0%) the patient's symptoms were recorded or were implied by the diagnosis. Information on the duration of symptoms was provided by the GP in 54 cases (10.8%). A past ocular history was provided in 50 cases (10.0%) and information on topical treatment already prescribed in 30 (6.0%). Some information on general health was included in 112 referrals (22.4%) and this proportion rose to 27.5% for those 127 patients complaining of loss of vision. A note of currently prescribed drugs was included in 67 cases (13.4%). For those 255 patients referred initially by the 00, an accompanying letter was sent by the GP in 74 (29.0%) cases in addition to the form GOS 18. In 132 referrals (51.8%) form GOS 18 alone was sent, with or without further information included. In 49 (19.2%)
instances a letter was sent but form GOS 18 was not enclosed. For those patients referred initially by the 00, the following additional information was supplied by the GP: ocular history 24 (9.4%); present ocular medication 2 (0.8%); past medical history 62 (24.3%); drug history 36 (14.1%). In 166 referrals (65.0%), no clinical patient information was supplied by the GP.
Discussion The spectrum ofdiagnoses shown in Table 1 illustrates patients' perceptions of the roles of GPs and opticians. It also reflects opticians' greater knowledge of ophthalmology and greater skill in ophthalmoscopy. Both professionals have an active part to play in the primary care of ocular problems. Ocular complaints account for 2.7% of all GP consultations12, yet undergraduate medical education in ophthalmology does not adequately reflect this. There is much room for improvement. Clearly, a referring doctor is selective about what is included in a referral letter. For this reason it should not be assumed that failure to include evidence of an ocular examination, indicates that no examination has taken place. However, for the ophthalmologist such information is important and the inclusion of normal findings is relevant. It is therefore disappointing that 43.7% of referrals from GPs contained no evidence of an ocular examination, and that only 6.1% of patients from this source had their optic discs or maculae examined. It is also of some concern that only 29% of patients referred by QOs included any comment on the posterior segment, despite the fact that examination of the optic discs is a statutory part of a refraction test. The measurement of visual acuity is, for the ophthalmologist and optician, the first part of any ophthalmic examination. This need not be the case for the general practitioner, but should be mandatory if the patient complains of visual loss or other visual symptoms. The test is not time-consuming and every general practice should be able to provide a Snellen chart for this purpose. It is of concern that even for patients with visual loss, only 6.3% of patients had visual acuity measured by their GP. The figure of 71.7% for opticians was surprisingly low. The importance of providing a medical background for a patient with visual loss was appreciated in only the minority ofreferrals (27.5%). In only 13.4% was any drug information provided. Information on currently prescribed drugs is of great importance in any referral, and no less so because the recipient is an ophthalmologist. Many ophthalmological patients are elderly, and it is in this age group where much inaccuracy in drug reporting is found7. Some drugs can cause ocular disease (eg prednisolone, chloroquine). Others can interact with topical ophthalmological drugs (eg Verapamil with ,8-blockers) whereas some can potentiate drugs used in ophthalmological practice (eg potassium-losing diuretics and aceta7nlamide). Some have important implications for surgery, and this includes minor eyelid surgery (eg warfarin). Drug information includes topical treatment, and in only 6% of this study was such information included. We are aware that by no means all patients will have used topical treatment, but it is common for such information to be absent when treatment has been given. The information is important especially where an antibiotic has already
771
772
Journal of the Royal Society of Medicine Volume 83 December 1990
been tried, and is crucial if a topical steroid has been prescribed. For those patients initially referred by their optician, 65% had no further information supplied before the referral was forwarded to the ophthalmologist. It was also disappointing that in 19.2%, form GOS 18 was not sent on by the GP. With this referral route, both optician and GP have impQrtant information to divulge. Form GOS 18 contains a refraction and details of an ophthalmic examination which will be comprehensible to an ophthalmologist but often not to a GP. This form should always be included. Similarly, many patients with ocular complaints will have underlying medical problems unknown to their optician, which can be placed into perspective by their general practitioner. This opportunity should not be missed. Suggestions that opticians refer directly to ophthalmologists should be resisted for this reason, except in the occasional purely ophthalmic emergency such as acute glaucoma. The question of screening for ocular diseases, especially glaucoma, has been raised in the past and the suggestion has been made that general practitioners have a role to play13. Opticians are however better placed to do this. They are better trained in ophthalmology (though this is a condemnation of undergraduate medical training), are more likely to be accurate in their diagnosis614"5 and have access to tonometry and visual field testing, which is necessary to make a better assessment of the patient than is possible by mere examination of the optic discs16. That opticians are already more active in this field is supported by this study, in which GPs referred two glaucoma suspects, and opticians 47. In an attempt to improve the quality and layout of information provided in a hospital discharge Practice stamp
Date
Dear
DOB
Re:
Address:
Occupation: Now Patient
Eye Hosp No: Symptoms:
Duration: Treatment:
Exasination:
Best Visual acuity
R
L
Diagnosis:
Hedical History:
Current drugs:
Priority:
Routine
Some priority
Urgent
Reason:
Yours
Figure 1. Proposed format for a standard referral letter
summary, a standard format has been suggested8. The concept can equally be applied to a referral letter, and an example is given in Figure 1. The letter contains practice and patient details, and is followed by a concise list of information which would enable rapid and accurate assessment of the patient's needs by the ophthalmologist. These details include some statement on symptoms and their duration, the essentials of ocular examination, and supporting information on past medical history and prescribed drugs. The widespread usage of such forms, together with the mutual development of referral protocols as suggested by Marinker et aL3 can only improve the standard ofreferral and therefore the service provided to patients. The provision of a similar section along these lines would also enhance the GOS 18 referral form. Acknowledgments: We are grateful for the willing help of the staff of the Medical Records Department, Manchester Royal Eye Hospital. References 1 Fair JF. Hospital discharge and death communications. Br J Hosp Med 1989;42:59-61 2 McGlade KJ, Bradley T, Murphy GJ, Lundy GP. Referrals to hospital by general practitioners: a study of compliance and communication. BMJ 1988;297: 1246-8 3 Marinker M, Wilkin D, Metcalfe DH. Referral to hospital: can we do better? BMJ 1988;297:461-4 4 Hull FM, Westerman RF. Referral to medical outpatients department at teaching hospitals in Birmingham and Amsterdam. BMJ 1986;293:311-14 5 Coulter A, Noone A, Goldacre M. General practitioners' referrals to specialist outpatient clinics. BMJ 1989:304-8 6 Harrison RJ, Wild JM, Hobley AJ. Referral patterns to an ophthalmic outpatient clinic by general practitioners and ophthalmic opticians and the role of these professionals in screening for ocular disease. BMJ 1988; 297:1162-7 7 Gilchrist WJ, Lee YC, Tam HC, MacDonald JB, Williams BO. Prospective study of drug reporting by general practitioners for an elderly population referred to a geriatric service. BMJ 1987;294:289-91 8 Penney TM. How to do it. Dictate a discharge sutmmary. BMJ 1989;28:1084-5 9 Penney TM. Delayed communication between hospitals and general practitioners: where does the problem lie? BMJ 1988;297:28-9 10 Mageean RJ. Study of "discharge communications" from hospital. BMJ 1986;293:1283-4 11 Gibson JM, Rosenthal AR, Lavery J. A study of'the prevalence of eye disease in the elderly in an English community. Trans Ophthalmol Soc UK 1985;104: 196-203 12 Dart JK. Eye disease at a community health centre. BMJ 1986;293:1477-80 13 Hitchings RA. Visual disability and the elderly. BMJ 1989;298:1126-7 14 Clearkin L, Harcourt B. Referral pattern of true and suspected glaucoma to an adult ophthalmic outpatient clinic. Trans Ophthalmol Soc UK 1983;103:284-7 15 Brittain GP, Austin DJ, Kelly SP. A prospective survey to determine sources and diagnostic accuracy of glaucoma referrals. Health Trendls 1988;20:43-4 16 Wood CM, Bosanquet RO. Limitations of direct ophthalmoscopy in screening for glaucoma. BMJ 1987;294: 1587-8
(Accepted 8 May 1990)
