Journal ofMedical Genetics, 1978, 15,435-442

A report on genetic Registers

Based on the Report of the Clinical Genetics Society Working Party1

ALAN E. H. EMERY (CHAIRMAN), C. BROUGH (COMMUNITY MEDICINE SPECIALIST),M. CRAWFURD (MEDICAL GENETICIST), P. HARPER (MEDICAL GENETICIST), R. HARRIS(MEDICAL GENETICIST), AND G. OAKSHOTT (GENERAL PRACTITIONER)

A Register may be defined as a record in whichregular entry is made of details of any kind sufficientlyimportant to be exactly recorded (Shorter OxfordDictionary). Genetic Registers range from simple cardindex systems of patients in whom a geneticist has aspecial interest, to elaborate computerised systemswhich may include a great deal of personal, medical,and genetic information on affected individuals andtheir relatives. There are a number of different types ofgenetic Registers, but our attention has been mainlydirected to Registers designed specifically for theprevention of genetic disease, since it is in this areathat most concern regarding feasibility, acceptability,and confidentiality has been voiced. In recent yearsthere has been growing interest in establishing suchRegisters as a means of improving the ascertainmentand follow-up of individuals in the population who areat risk of transmitting a serious genetic disorder totheir offspring.

In 1977 the Committee of the Clinical GeneticsSociety formed a Working Party in order specifically:

(a) To consider the nature of genetic Registers.(b) To consider the value and potential application

of such Registers, particularly with regard tothe prevention of genetic disease.

(c) To consider genetic Registers with regard to: (i)specific genetic disorders such as Huntington'schorea; (ii) genetic disorders in general whererelatives may be at high risk.

(d) To consider both the feasibility and accep-tability of genetic Register systems organised:(i) on a national basis; and (ii) on a more localbasis.

(e) To make recommendations.

'Dr Pauline Robertson acted as Secretary.The views expressed in this discussion document are those of the WorkingParty and not necessarily those of a}l members of the Clinical GeneticsSociety.Received for publication 12 August 1978

The Working Party met on several occasions todiscuss these matters and this paper represents asummary of some of the main conclusions in theReport, which was presented to the Society in July1978.

The nature and purpose of genetic Registers

At present, 38 centres in Britain maintain Registersprimarily concerned with genetic disease. TheseRegisters can conveniently be considered under fivedifferent headings, though they are not mutuallyexclusive.

(1) Clinical (or Therapeutic) Registers, to facilitatethe follow-up and recall of individuals for therapeuticreasons, or when relevant preclinical or prenataldiagnostic tests become available.

(2) Reference Registers, to store clinical andlaboratory data on individuals with genetic disease, sothat the diagnosis in a new case might be confirmed byreference to previous cases.

(3) Registersfor Monitoring Services, to assess theresults of genetic counselling and prenatal diagnosis.

(4) Research Registers, to store and analyseepidemiological data, or to determine the naturalhistory of particular genetic disorders.

(5) Preventive Registers, to improve the ascertain-ment and follow-up of individuals at risk of trans-mitting a serious genetic disorder to their offspring, sothat they may be offered genetic counselling andprenatal diagnosis where possible.Of the 90 Registers currently maintained in Britain,

11 would be considered Clinical/Therapeutic, 15Reference, 23 Monitoring, 29 Research, and 12Preventive.

Registers and the prevention of genetic disease

The results of several theoretical studies have shownthat unifactorial disorders, especially autosomal domi-

435

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

Emery, Brough, Crawfurd, Harper, Harris, and Oakshott

nant and X-linked recessive disorders, provide the bestscope for preventing genetic disease (Smith, 1970;Holloway and Smith, 1973; Fraser, 1974). Follow-up studies of individuals referred for genetic counsellinghave confirmed these theoretical findings, but havealso shown that over 85% of relatives deemed to beat high risk (greater than 1 in 10) of having a childwith a serious genetic disorder have never been givencounselling and are unaware of the risks (Emery andSmith, 1970; Emery, 1972). Further, a number ofaffected children were born to parents who, a priori,were at high risk of having an affected child, whichmight have been prevented if the parents had hadgenetic counselling. The ascertainment and follow-up of individuals at risk in the population is, there-fore, crucial in the prevention of genetic disease, andit is in this regard that a genetic Register system canbe of particular value, especially for the follow-upof those who are currently too young for geneticcounselling. A WHO Scientific Group (WorldHealth Organization, 1972) has, in fact, recommendedthat medical genetics centres should set up Registersof genetically determined disorders specifically forthe purpose of preventing genetic disease.

Particular consideration was given to three com-puterised genetic Register systems designed, at least inpart, for the purpose of preventing genetic disease: theEdinburgh RAPID system, the Indiana MEGADATsystem, and the Belgian National Register.

(1) THE EDINBURGH RAPID SYSTEMA Register system specifically designed as a means ofascertaining and preventing genetic disease was firstestablished in Edinburgh in 1970 under the acronymRAPID (Register for the Ascertainment and Preven-tion of Inherited Disease), and has been fully opera-tional for the last two years (Emery et al., 1974;Emery, 1976). The organisation of the RAPID systemcan be conveniently considered under two mainheadings: ascertainment, contact, and follow-up ofindividuals at risk; and the recording, storage, andretrieval of family data.

(a) Ascertainment, contact, andfollow-upAscertainment of individuals at risk of having a childwith a serious genetic disorder may be direct or in-direct. Direct ascertainment is when the index case isbrought to attention as a result of routine diagnosis.Indirect ascertainment is when index cases areobtained through public health records and statutoryregisters, and therefore without the individual's know-ledge. Indirect ascertainment raises a number ofdifficult ethical problems, and is also extremely time-consuming and relatively unproductive. This form ofascertainment has therefore been abandoned, andindividuals at risk are now only ascertained directly

through special clinics, which include a large propor-tion of inherited disorders (for example, neurology,haematology, and ophthalmology), special schools(for example, for the mentally handicapped, blind, andprofoundly deaf), and various voluntary patientorganisations.

Having ascertained individuals at risk of havingaffected children, at present only those at high risk(greater than 1 in 10) are contacted and followed-up.Relatives at risk are contacted only with the expresspermission of both the index case and the relative'sfamily doctor, and, at least in the first instance,contact is best arranged directly by the familydoctor or health visitor. This is considered important,as there may be factors unknown to the geneticistor index case which might make it unnecessary, oreven imprudent, to contact certain family members.

At present, any individual has the prerogative towithhold knowledge of his medical condition from hisrelatives if he so wishes, though in taking this attitudehe accepts a heavy responsibility if such informationmight prevent the birth of an affected child in anotherbranch of the family. No individuals are included inthe RAPID system, nor are their relatives contacted,without written permission, a specially designed formbeing used for this purpose. This form is signed onlyafter the Register system has been explained and theimplications discussed with the individual concerned.

(b) Recording, storage, and retrieval offamily dataFor ease of storage and retrieval of family data, theRAPID system has been computerised. The presentfile system can store data on 25 000 to 30 000 indivi-duals. Access to the data is through an interactive tele-type terminal using the Edinburgh Multi-AccessSystem (EMAS) on an ICL 4-75 computer. Thecomputer program is in FORTRAN, details of whichare available (Moores and Emery, 1976). To main-tain strict confidentiality of the information in theRegister, a number of security checks have been incor-porated into the system and the various recommenda-tions of the Younger Report have been adopted(Computers and Privacy, 1975).

Currently, information on over 700 individuals isstored in the RAPID system, involving 52 differentdisorders, though about 80% of those on the Registercan be accounted for by relatively few disorders(Table 1).

(2) THE INDIANA MEGADAT SYSTEMMEGADATS (Medical Genetics Acquisition andData Transmission System) is a computerised systemfor the collection and processing of information onfamilies with genetic disorders. It was set up by theIndiana Human Genetics Centre using information on

436

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

A report on genetic Registers

Table 1 Proportions (%) of individuals at risk on RAPID

Disorder % ofindividuals

Autosomal dominantsHuntington's chorea 15Polycystic kidney disease 9Myotonic dystrophy 5Neurofibromatosis 3Polyposis col 2Retinitis pigmentosa 2Marfan's syndrome 2

X-linked recessivesHaemophilia (A and B) 27Duchenne muscular dystrophy 7Retinitis pigmentosa 3

Familial translocationDown's syndrome 3

families seen for genetic diagnosis, counselling, or forresearch studies, and in 1975 it incorporated data on a

third of a million individuals from over 15 000families, and is growing by about 1500 families per

year (Merritt et al., 1976).MEGADATS was set up to facilitate the storage,

updating, and processing of data collected for use inmedical genetic services as well as for research. Theaims have been set out as being:

(1) To facilitate the diagnosis of inherited disorders.(2) To provide greater insight into the genetics of

complex problems with major genetic compo-

nents, and thereby aid genetic counselling, earlydiagnosis, and preventive therapy.

(3) To show the value of these data to practi-tioners through ready access, cost-effective-ness, and patient response to counselling.

(4) To encourage the use of pooled data fromspecialised populations with a high risk ofcertain specified diseases, both simply inheritedand of unknown complex aetiology.

(5) To show the feasibility of collecting, recording,and using this highly personal and sensitivedata while fully protecting confidentiality andindividual rights.

(6) To permit the retrieval of information on

heterogeneity, genetic linkage, and chromo-some mapping that will assist in diagnosis andcounselling by providing improved risk estim-ates.

The system has been built up over the last 14 years.

Initially, medical genetics department files were storedon magnetic tape and updated at 6 to 12 monthintervals. They were processed on an IBM 7040 com-

puter in the Research Computation Centre of theUniversity Medical Certre by batch processing. Thedata were used for biometrical studies, statistical, andgenetic analyses.By 1975, a PDP-11/45, coupled to a PDP-11/05

for laboratory automation, was also in use. This is

437

used interactively and includes, among a wide varietyof facilities, an XY plotter and XY digitiser. It is linkedto the Indiana University regional computer network.In order to retrieve information in an ordered manner,a data base management system (S2K) is used, whichfacilitates retrieval of pedigree material. Security ismaintained by password control to the data base andto each component.MEGADATS is thus a complex computer system

for the storage and retrieval of a variety of geneticinformation on individual families. Such informationcan be used for both clinical and research purposes.

(3) THE BELGIAN NATIONAL REGISTERThis is a Register containing data on all personsreferred to any of the seven Belgian medical geneticcentres, which amounts to about 5000 to 6000 peoplea year (Vlietinck and Van den Berghe, 1976). There isa master file containing a unique identification of eachindividual, but this is available only to the referringcentre and is otherwise used only to ensure that a newentry is not already on the file. Other data, such as sex,birth place, consanguinity, non-paternity, adoption,mode of ascertainment, ethnic origin, pedigree, andmedical information, are kept in further files and areavailable to all the participating centres. Healthinsurance companies in Belgium maintain Registers ofindividuals with malformations, cancer, tuberculosis,and mental illness. There is no link between these andthe genetic Register, but the former can provideepidemiological data and, in the case of one of thelarger Registers, records its data in a similar manner tothe genetic Register. Other programs cater forpedigree analysis, linkage, recurrence risk calcula-tions, and statistical analyses. Thus, this Register isvery similar in many ways to MEGADATS. However,unlike MEGADATS, it is a national Register andshould therefore be able to provide epidemiologicaldata relating to the entire population.

Registers for specific genetic disorders and for geneticdisorders in general

A number of Registers for specific genetic disordersalready exist, for example, Registers for haemophilia(several centres), polyposis coli (as at St. Mark'sHospital, London), and phenylketonuria (as at theInstitute of Child Health, London). Other genetic dis-orders for which Registers already exist, or are beingconsidered, include Huntington's chorea, polycystickidney disease, myotonic dystrophy, retinitis pigmen-tosa, and Duchenne muscular dystrophy.

Registers for specific genetic disorders have certainadvantages over more general genetic Registers. Thereis likely to be increased accuracy of diagnosis andbetter quality of information, greater familiarity of the

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

Emery, Brough, Crawfurd, Harper, Harris, and OakshottUnit maintaining the Register with problems resultingfrom the disorder, more complete ascertainment in aregion1, and the possibility of closer integrationbetween genetic matters and the general managementof the disorder.

However, against these advantages must be set thedisadvantages of fragmentation resulting from differ-ent Registers being kept by different departments, andlack of standardisation, since the form of a Registermay differ widely depending on the nature of thedisease. In setting up a specific disease Register, it isimportant that the system of data recording be com-patible with more general Registers, as well as withother specific disease Registers.

Feasibility of genetic Registers

Although several genetic Registers have now beenestablished, there have been very few feasibility studiespublished. One reason for this probably lies in therather imprecise aims of many of the Registersestablished so far.

CO ST/BENEFITCost/benefit analysis is notoriously difficult inmedicine, and perhaps especially so in the case of suchproblems as genetic Registers. Emery estimated in1974 that the overall cost of RAPID, exclusive ofsalaries, was £1*L09 per individual counselled. By 1977this figure had risen to £3.50 per individual. Thesefigures include the costs of correspondence and travelinvolved in family visiting, as well as computing, andwere higher for consultands2 than for their relatives.Thus, costs for any other Register would depend onthe ratio of these two categories of individuals coun-selled. To these administrative and computer costsmust be added the salaries of a medical geneticist, aclerk/computer operator, and a field worker. This isthe staff considered necessary to maintain a Registerfor a population of a million, which might generateapproximately 200 new families each year. Extensionto a larger population would require additional resour-ces in regard to field workers and medical geneticists.

Benefit is far more difficult to estimate in financialterms, since it depends on how many affected birthswill be prevented through genetic counselling, and alsoon the wide variation in cost to the community ofcaring for patients with different serious genetic dis-orders.Rough estimates of the number of cases of some

genetic disorders which might be prevented by geneticcounselling in the United Kingdom each year are given'In this Report, region is used as a term of convenience and is not necessarilymeant to designate a specific administrative area within the NHS.2A consultand is defined as the first member of a family who seeks geneticcounselling. He may be healthy, affected, a preclinical case, or a carrier.

Table 2 Approximate numbers of cases of some geneticdisorders which might be prevented by genetic counsellingin the United Kingdom each year (Emery, 1978).Disorder Birthfrequency Estimatedpreventable

(per 1000 live births) cases each year

Autosomal dominantsHuntington's chorea 0.50 350Polycystic kidney 0.80 448diseaseMyotonic dystrophy 0-20 126Neurofibromatosis 0-40 140Polyposis coil 0.10 49Marfan's syndrome 0.04 22

X-linked recessivesHaemophilia 0.10 29(A and B)Duchenne muscular 0-30 22dystrophy

It is assumed that there are 700 000 live births (360 000 males) per yearin the UK (based on data for 1975). Birth frequencies are based onvarious sources, but mainly Carter (1977), and in X-linked recessivedisorders this refers to the frequency per 1000 live male births.

in Table 2, but these do not include any benefits thatmight accrue from population screening programmes.Since many of these disorders will cost the com-munity, in medical and welfare services, considerablymore than the expense of ascertainment and counsel-ling, even if the latter is only partly effective it wouldseem likely that this approach could be well justifiedon economic grounds alone.

RESPONSIBILITY FOR GENETIC REGISTERSA genetic Register covering a particular region wouldbe most conveniently based in the principal medicalgenetic centre for that region. Because of the clinicalnature of much of the data stored, and the probabilitythat most genetic Registers that will be set up in thiscountry will be for the purpose of improving ascer-tainment for genetic counselling, the appropriateindividual to have overall responsibility for theRegister would be a medical geneticist of consultantstatus, with the co-operation of appropriate commun-ity medicine staff.

ACCESS TO GENETIC REGISTERSThe question of access depends on the aims of theRegister. A Register maintained purely for ascertain-ment and follow-up of individuals at risk, so that theycan be offered genetic counselling, should only allowaccess to persons actually involved in genetic coun-selling. Persons able to retrieve information from theRegister will be more narrowly defined than those whoare allowed to enter information. As individuals at riskwill not be confined to a particular health serviceregion, it is essential to provide a service for geneticcounsellors outside the region.

Registers with broader functions, such as the main-tenance of epidemiological or research data, will needto provide for wider access. This may require a more

438

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

A report on genetic Registerscomplex categorisation of types of worker grantedaccess in relation to levels of information within theRegister. For example, epidemiologists and mostresearch workers may be granted access to statisticaldata but not to personal data. If individuals are to beidentified to research workers, this would require themost stringent safeguards and the fully informedconsent of the individuals concerned, or their guar-dians, on each separate occasion that the informationwas released. Probably the only really adequate safe-guard for this type of access is that it should not bedirect, but only through the centre in charge of theRegister.

AVAILABILITY OF SUPPORTBefore initiating a genetic Register, it is essential toestablish that adequate provision has been made forgenetic counselling services, and for medical and socialsupportive help for those who are found to be at risk.The establishment of a Register will generate a workload which may well- exceed the existing geneticservices. Further, those ascertained as being at risk ofhaving affected children may require additional sup-port from the medical profession as well as the socialservices.

Acceptability of genetic Registers

The acceptability of genetic Registers to those mostlikely to be involved, namely consultands and theirrelatives and family doctors, is an important issue andwas considered in some detail.

ACCEPTABILITY TO CONSULTANDS ANDRELATIVESNumerically, the most important groups of individualslikely to be concerned with acceptability are those whoare already affected with a genetic disorder or are pre-clinical cases, and those who are carriers of X-linkedrecessive disorders. Wilkie and Sinclair (1977) studiedacceptability in regard to RAPID and concluded thatthose involved with X-linked recessive disorders werethose who were most in favour of genetic Registers.However, the advantages of counselling an individualat risk of developing a disorder such as Huntington'schorea, myotonic dystrophy, or neurofibromatosismay not always be evident to the individual con-cerned. The obvious advantage is that it may helpprevent the disorder from being transmitted to subse-quent generations. The main disadvantage is that sincean effective treatment is not yet available for any ofthese disorders the knowledge that one may developthe disease may engender considerable distress. Suchinformation may in some instances lead to marital dis-harmony, but the alternative of withholding informa-tion or erroneously reassuring a couple is to be

439deprecated. An important function of a geneticRegister should be to help detect individuals at riskbefore marriage, so that the individual would be in aposition to inform a possible spouse of the problemand its implications. An entirely satisfactory solution,however, must await the discovery of effective treat-ments for these disorders and/or reliable precinicaland prenatal diagnostic tests.Of course, it is important to realise that through a

genetic Register system, relatives who prove to be atlow risk can be reassured and anxieties, which oftenexist in such family members, can be dispelled.

Attitudes to the acceptability of genetic Registerswill not only vary from disorder to disorder, but willalso depend on the amount of support, both voluntaryand statutory, which individuals may have experien-ced. The family doctor may be able to play a majorpart in the support of families involved with geneticdisorders. Wilkie and Sinclair (1977) found that over95% of family doctors in the south-east of Scotlandwere in favour of a genetic Register, but wheneverpossible they felt they would like to be involved in thecounselling and follow-up support, though they lookedto the geneticist for information about diagnosis andrisks.

ACCEPTABILITY TO THE FAMILY DOCTORA very important problem with any genetic Registersystem is its acceptability to family doctors.' Animportant function of the family doctor, in the light ofhis intimate knowledge of the family situation, is toaugment the information given to a patient by thegeneticist, and to ensure that continuing support forthe patient and his family is available when required(World Health Organization, 1969). Satisfactory caremay involve a continuous exchange of information, onthe patient's behalf, with non-genetic specialists andwith a variety of non-medical agencies. These mayinclude the DHSS (for example, invalidity allowance),educational and housing authorities, and telephoneservices. In this dissemination of information, there area number of constraints imposed upon the attitude ofthe family doctor by the nature of genetic disease.They include concern for confidentiality, invasion ofprivacy, and the patient's loss of self-esteem. Thesematters were considered in detail.

It should be emphasised that the family doctormight reasonably be expected to play an importantrole in the operation of any genetic Register. Becauseof his often intimate knowledge of the family situation,he is not only in the best situation to protect hispatients from any possible unnecessary intrusion, butin many situations he may well be able to give geneticcounselling himself, and certainly provide support andhelp for those found to be at high risk.

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

0Emery, Brough, Crawfurd, Harper, Harris, and OakshottGenetic Registers: national or regional?

Advantages and disadvantages exist for both nationaland regional genetic Registers, and the balance islikely to depend both on the disorder being consideredand the aims of a particular Register.The advantages of a national Register include the

increased amount of data compared with that recor-ded by a local Register. This may make registration ofa very rare disorder worthwhile, and may providesufficient data on a less rare disorder to permitmeaningful epidemiological analysis. A nationalRegister also allows standardisation of data collectionover the whole country, overcomes the increasingproblem of geographical dispersal of relatives, andfacilitates adequate staffing and funding from centralsources without unnecessary duplication. However,disadvantages include the difficulties of ensuring thatdata are complete and accurate. A lack of personalknowledge of the families registered by those running anational Register is a particularly important factor, aswell as the increased difficulties of trying to maintainconfidentiality and restriction of access to valid users.This could well reduce the co-operation of familieswho might otherwise be willing to be included on amore local Register.

Taking into account these various considerations, itwould seem probable that, at least in Britain, a nationalRegister of genetic -disorders in general is likely to beundesirable, and that the national level is also not suit-able for Registers for specific genetic disorders, theaim of which is the identification and recall ofindividuals at risk. A national Register could, however,be appropriate for certain rare disorders for whichinformation would be too limited at a local level, andwhere the aim is primarily to study the natural historyand management over a prolQnged period, but suchRegisters would be the exception.

Use ofcomputers for genetic Registers

Computerisation of data in a genetic Register becomesimportant for ease of storage and retrieval when largenumbers are involved, but otherwise a manual systemcan be quite adequate. There are, however, severalother advantages of using a computer. When appro-priate safeguards have been incorporated into thesystem, it helps maintain and enhance confidentiality.It also provides a facility for including other functions,such as programs for risk calculation (Smith, 1972;Conneally and Heuch, 1974).

With regard to computer specifications, these willdepend largely on the size of the population to becovered. For a population of a million, it might beexpected that 10 000 (that is, about 1%) would bd an

upper limit to the number of individuals (patients andrelatives) who might be included on a genetic Register.With RAPID, each individual has a total record ofabout 500 characters and, therefore, on this basis aRegister of 10 000 would take up to 5 million charac-ters. This is equivalent to about 2 RK05 on a PDP 11,and in current terms is not excessive. With regard tothe amount of core store required, this is more difficultto estimate, as it will depend on the operating systemand the computer language chosen, but 16 K wordswould be a minimum. Clearly with more core avail-able, the programmer's task will be easier.

Another problem to be considered is the form thatdata input and output should take. With regard toinput, a key system is probably the most practical, andsince a written record is often required, some form ofdata printout might well be preferred.

Finally, there is the type of computer facility to beconsidered. There are a number of distinct advantagesassociated with a small dedicated computer actuallylocated in the department of the main users, asopposed to a terminal associated with a large centralcomputer. These advantages include increased con-fidentiality (prevention of unauthorised use and betterphysical security) and facility of operation (avail-ability of interactive mode and immediate access fordesignated users).

Whatever system is chosen, there would be obviousadvantages in using a standard form for recordingdata and an agreed system for coding individual itemsof information, such as diagnosis. This would helpfacilitate the exchange of data between differentgroups, and might hopefully lead to a standard methodof data storage in the computer file. In some circum-stances, it may even be possible for several groups touse the same program and hence reduce develop-ment costs. Unfortunately, there is as yet no generallyagreed classification of disease which would be ofspecific value in a genetic Register. The 'InternationalClassification of Disease' is not sufficiently detailed inregard to genetic syndromes and disorders. The'Cardiff Diagnostic Classification', produced by theBritish Paediatric Association, has the advantage ofbeing more discriminating in the case of certain geneticdisorders because of the addition of a 5th and some-times a 6th digit. However, an important disadvantageis that ICD codes are not always adhered to. Thecoding system of genetic disorders given by McKusick(1975) is useful, but unfortunately does not includemultifactorial or chromosomal disorders.A suggested coding, based on the 'International

Classification of Disease' to which a 6th digit has beenadded to accommodate increasing heterogeneity withinparticular disease entities, has been devised andemployed in the RAPID system. This coding systemaccommodates various unifactorial, multifactorial,

440

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

A report on genetic Registersand chromosomal disorders, and details are availableon request'.

Finally, but most importantly, consideration has tobe given to data security. This can be achieved inseveral ways. Firstly, by encoding data which mightidentify individuals in the Register. Secondly, if thedata are stored on disk or cassette, these should beremoved from the machine and locked away when notin use. Thirdly, the system should be so designed thatthe data can always be recovered, even in the event ofa computer malfunction, by having back up copies ofthe data which are stored securely.

Recommendations

As a result of their deliberations and consideration ofthe evidence, the Working Party proposed the follow-ing recommendations:

(1) Registers should be set up for the expresspurpose of tracing, following-up, and counsellingindividuals who are at high risk (greater than 1 in 10)of transmitting a serious genetic disorder to their off-spring. The object is to ensure that such individualsare informed of the risks and the various options avail-able to them.

(2) Registers should be organised on a regionalrather than a national basis, and should be located in aregional genetics centre. When Registers already existin Units dealing with specific genetic disorders, such ashaemophilia, close liaison should be encouraged withthe regional genetics centre in order to providecomprehensive facilities for counselling, including riskestimation.

(3) No individuals should be included on theRegister without their full knowledge and approval.

(4) The involvement of family doctors should besought in every case and encouraged at all stages ofcontacting, counselling, and follow-up of individuals atrisk.

(5) The complexity of the data and the need for co-operation between centres requires a standard form forrecording data and an agreed system for codingdisease entities. The latter should accommodatechromosomal, multifactorial, and unifactorial dis-orders and should be sufficiently flexible to includenewly recognised genetic disorders and possibleheterogeneity.

(6) Computerisation of data in a genetic Register isrecommended for several reasons. It becomes impor-tant for ease of storage and retrieval of data (includingpatient recall) when large numbers are involved. It hasthe advantage that, when appropriate safeguards areincorporated, it helps to maintain strict confi-

'From Professor A. E. H. Emery, University Department of Human Genetics,Western General Hospital, Edinburgh EH4 2XU.

441dentiality. It also provides a facility for fulfilling otherfunctions such as complex risk calculations.

(7) The preferred computer system for a geneticRegister requires a dedicated mini-computer located inthe genetic centre. The most important advantages areincreased confidentiality and facility of operation. Aless satisfactory alternative is a terminal associatedwith a large central computer.

(8) There should be standardisation of computerequipment and computer programs, both for eco-nomic reasons and to facilitate collaboration betweendifferent genetic centres. Therefore, it is essential to co-ordinate developments in this field.

(9) The computer program, like all other aspectsof the genetic Register system, must incorporate strictsafeguards for confidentiality.

(10) Access to data in the Register must berestricted to certain clearly specified individuals.Personal and medical information should be releasedonly by the clinician in charge of the Register.

(11) In establishing a regional genetic Register, pro-vision must be made for adequate supporting staff,which might reasonably include a medical geneticist, aclerk/computer operator, and a field worker.

(12) Where a general genetic Register is notimmediately feasible for economic or other reasons, alimited Register should he established for specificgenetic disorders of particular importance.

We are very grateful to Mrs P. Wilkie and Mr M.Moores, who presented both written and verbal evi-dence at two of our meetings, and to all those who verykindly responded to our request for information aboutthe various types of Registers which they currentlymaintain.

ReferencesCarter, C. 0. (1977). Monogenic disorders. Journal of Medical

Genetics, 14,316-320.Computers and Privacy (1975). Cmnd. 6353. HMSO, London.Conneally, P. M., and Heuch, I. (1974). A computer program to

determine genetic risks: a simplified version of PEDIG. AmericanJournal ofHuman Genetics, 26,773-775.

Emery, A. E. H. (1972). The prevention of genetic disease in thepopulation. International Journal of Environmental Studies, 3,37-41.

Emery, A. E. H. (1976). RAPID: a genetic register system for theascertainment and prevention of inherited disease. In Registersfor the Detection and Prevention of Genetic Disease, pp. 53-63.Ed. by A. E. H. Emery and J. R. Miller. Symposia Specialists,Stratton Intercontinental, New York.

Emery, A. E. H. (1978). Prevention of genetic disease. In AdvancedMedicine, Vol. 14, pp. 118-128. Ed. by D. J. Weatherall. PitmanMedical, London.

Emery, A. E. H., Elliott, D., Moores, M., and Smith, C. (1974). Agenetic register system (RAPID). Journal of Medical Genetics,11, 145-151.

Emery, A. E. H., and Smith, C. (1970). Ascertainment andprevention of genetic disease. British Medical Journal, 3, 636-637.

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from

Emery, Brough, Crawfurd, Harper, Harris, and Oakshott

Fraser, G. R. (1974). The long-term genetical effects of recentadvances in the treatment and prevention of inherited disease.British Journal ofPsychiatry, 125, 521-528.

Holloway, S. M., and Smith, C. (1973). Equilibrium frequencies inX-linked recessive disease. American Journal ofHuman Genetics,25,388-396.

McKusick, V. A. (1975). Mendelian Inheritance in Man, 4th ed.Johns Hopkins University Press, Baltimore and London.

Merritt, A. D., Kang, K. W., Conneally, P. M., Gersting, J. M., andRigo, T. (1976). MEGADATS: a computer system for familydata acquisition, storage and analysis. In Registers for theDetection and Prevention of Genetic Disease, pp. 31-51. Ed. byA. E. H. Emery and J. R. Miller. Symposia Specialists, StrattonIntercontinental, New York.

Moores, H. M., and Emery, A. E. H. (1976). RAPID: a descriptionof RAPID, an information, storage and retrieval systemdeveloped for work in genetic registers. Inter-University ResearchCouncils Report Series. Edinburgh Regional Computing Centre,Edinburgh.

Smith, C. (1970). Ascertaining those at risk in the prevention andtreatment of genetic disease. In Modern Trends in HumanGenetics, Vol. I, pp. 350-369. Ed. by A. E. H. Emery. Butter-worths, London.

Smith, C. (1972). Computer programme to estimate recurrence risksfor multifactorial familial disease. British Medical Journal, 1,495-497.

Vlietinck, R. F., and Van den Berghe, H. (1976). The BelgianNational Genetic Register. In Registers for the Detection andPrevention of Genetic Disease, pp. 65-71. Ed. by A. E. H. Emeryand J. R. Miller. Symposia Specialists, Stratton Intercontinental,New York.

Wilkie, P., and Sinclair, S. (1977). Report of the Genetic RegisterSystem Acceptability Study. Department of Social Ad-ministration, University of Edinburgh.

World Health Organization (1969). Genetic Counselling. TechnicalReport Series, no. 416. WHO, Geneva.

World Health Organization (1972). Genetic Disorders: Prevention,Treatment and Rehabilitation. Technical Report Series, no. 497.WHO, Geneva.

Requests for reprints to Professor A. E. H. Emery,University Department of Human Genetics, WesternGeneral Hospital, Edinburgh EH4 2XU.

442

copyright. on S

eptember 6, 2021 by guest. P

rotected byhttp://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.15.6.435 on 1 Decem

ber 1978. Dow

nloaded from