initial response has been encouraging; however itis possible that some people who are working inthis field and ought to be approached have beeninadvertently missed.

We are therefore seeking the help of all re-searchers to ensure that a registration form iscompleted by the principal investigator for allrandomized controlled trials of a smoking cess-ation intervention of which they may be aware,and which is currently in progress or substan-tially advanced in the planning stages. To beeligible for inclusion in the registry, a trial must:(a) be unpublished; (b) include at least twogroups; (c) allocation to the groups must be byeither a random or quasi-random method (e.g.alternation, year of birth, etc.); and (d) the trialmust be related to an aspect of smoking cess-ation. Trials examining abstinence rates, relapseprevention, withdrawal symptoms, training orencouraging health professionals in smokingcessation techniques, or any aspect of smokingcessation research are all eligible for inclusion. Ifyou are in doubt as to whether a trial is suitablefor inclusion, we suggest you still complete aregistration form.

Once the register has been assembled a copywill be distributed to all contributors, as well aspublished in summary form on an annual orbiannual basis.

The registry will not collect any trial resultdata or participant information, although theexistence of such a register may facilitate effortsto establish collaborative groups who wish toundertake more detailed systematic reviews inthe future, similar to those undertaken in otherfields. Trial registration forms are available onrequest firom:

The Collaborative Registry of Smoking CessationTrials^ General Practice Research Group, GibsonBuilding, Radcliffe Infirmary, Oxford OX2 6HE,United Kingdom. Tel: +44-865-319 111 Fax:+ 44-865-511 635.

The coordinators would also appreciate beinginformed of any completed but unpublished smok-ing cessation trials of which you may be aware.No special form is provided for this purpose;however, any information that researchers canprovide will assist in up-dating our current regis-ter of completed trials and ensure its comprehen-siveness.

As the success of the registry will depend onthe amount of information provided by re-

Letters to the Editor 483

searchers, we would be grateful if you woulddraw the attention of your readers to thisregistry.

CHRIS SILAGY & GODFREY FOXXO^R

TTie Collaborative Registry of Smoking CessationTrials, General Practice Research Group,Gibson Building, Radcliffe Infirmary,Oxford OX2 6HE, UK

Reference1. CHALMERS, L, DICKERSIN, K. & CHAUVIERS, T . C.

(1992) Getting to grips with Archie Cochran'sagenda, British Medical Journal, 305, pp. 786-788.

Controlled drinking researchSIR—In May 1993 Professor Edwards gave a talkat the Addiction Research Foundation focusingon the impact of D. L. Davies's 1962 article onthe alcohol field, and in particular on a cohort ofpsychologists who conducted controlled drinkingresearch in the 1970s. In reacting to Edwards'sthesis, our view was that Davies's article was butone of several factors precipitating controlleddrinking research. Since little information isavailable on the history of controlled drinkingresearch, we thought it might be helpful to com-ment on the role we perceived Davies's article tohave played in our research.

A literature review that we conducted in 1969identified about 23 studies questioning an absol-ute abstinence requirement, and none demon-strating that controlled drinking was impossible.We also found several experimental studiesdemonstrating the malleability of the drinking ofchronic alcoholics in a controlled environment.Having come from a strong scientific back-ground and lacking indoctrination in traditionaltreatments, an objective appraisal of the evidencesuggested that controlled drinking was an openissue.

As a report of controlled drinking, Davies'sarticle was just one of several that appeared inthe literature and that infiuenced our early re-search. Judged by today's standards, few of thesestudies would be considered rigorous. However,by 1969 standards the studies as a whole were asmethodologically sound as other treatment out-come studies in the literature and in some caseswere more sophisticated. This is important, be-cause it is easy to lose perspective whenrefiecting on research reported a quarter of acentury ago.

484 Letters to the Editor

While we believe we would have conductedthe controlled drinking research whether or notDavies's article had been published, it was thereactions to Davies's article (Davies, 1963a, b)that most sparked our interest. The acrimoniousand often illogical responses to Davies's articlemade it clear that the question of controlleddrinking was controversial but unresolved. It isimportant to note that there was no publishedevidence showing that controlled drinking wasimpossible. Also, other articles significantlyinfiuenced our thinking, A few well written cri-tiques of the then contemporary thinking, suchas Pattison's 1966 paper, and Moore's 1962article, compellingly argued for conducting re-search on alternatives to abstinence.

Finally, it is important to note the influencethat behaviour therapy had on the alcohol field.In the 1960s, two eminent behaviorists, ArnoldLazarus (1965) and Albert Bandura (1969), hadput forth a conceptualization of alcohol prob-lems that naturally led to a consideration of goalsother than abstinence. Neither Lazarus nor Ban-dura seems to have been affected by Davies'swork. Rather, they were simply applying a learn-ing theory paradigm to the explanation of behav-ior. Their writings had a very substantialinfluence on our research.

In summary, Davies's 1962 study, while im-portant, was just one of several key factors that atthe time combined to suggest that research onalternatives to abstinence should be conducted,

MARK B, SOBELL & LINDA C, SOBELL

Clinical Research and Treatment Institute,Addiction Research Foundation,33 Russell Street,Toronto, Ontario, Canada, M5S 2S1

ReferencesBANDURA, A, (1969) Principles of behavior modification

(New York, Holt, Rinehart & Winston).DAVIES, D , L, (1962) Normal drinking in recovered

alcohol addicts. Quarterly Journal of Studies on Al-cohol, 23, pp, 94-104,

DAVIES, D , L, (1963a) Comment by various correspon-dents: normal drinking in recovered alcohol addictsby D, L, Davies (1962), Quarterly Journal of Studieson Alcohol, 24, pp, 109-121, 321-332,

DAVIES, D , L, (1963b) Comments by various corre-spondents on normal drinking in recovered alcoholaddicts by D, L, Davies (1962), Quarterly Journal ofStudies on Alcohol, 24, pp, 727-735,

LAZARUS, A, A, (1965) Towards the understanding andeffective treatment of alcoholism. South AfricanMedical Journal, 39, pp, 736-741,

MOORE, R, A, (1962) The problem of abstinence bythe patient as a requisite for the psychotherapy ofalcoholism: I, The need for abstinence by the al-coholic patient during treatment. Quarterly Journal ofStudies on Alcohol, 23, pp, 105-111,

PATTISON, E, M , (1966) A critique of alcoholism treat-ment concepts: with special reference to abstinence.Quarterly Journal of Studies on Alcohol, 27, pp, 49-71,

Beneficial effects of nicotine: fact or fiction?SIR—I read with interest Dr Robert West's arti-cle {Addiction, 1993, 88, pp, 589-590) in whichhe posits that: smokers may be constitutionallyworse off than non-smokers, and use cigarettesto increase their psychological functioning to apoint comparable with non-smokers. Comment-ing on the current debate on the psychopharma-cology of nicotine. West claims that smokers areno better off in cognitive tasks once social class,age and sex are controlled for, I believe he mayhave an interesting point. Research I have beeninvolved with (Baker & Walsh, 1993) using a(predominantly female) university sample 20 to27 years of age found no eflect of nicotine onvisual discrimination. Beneficial effects of nic-otine have not been entirely consistent, in re-search with both animal and human subjects (seeLevin, 1992), Given also that smokers are aheterogeneous group, an explanation for thevariable extent of nicotine addiction and its in-consistent cognitive effects could lie in terms ofcentral nervous system (CNS) 'tone'. It is appar-ent from patients with migraine or epilepsy thatthe constitution of the CNS can have profoundcognitive and behavioural effects, I believe thatto follow up hypotheses such as Dr West's, to-gether with more sophisticated experimentalcontrols, would make nicotine research moretheory-led than data-led, and possibly help settlean important and at times heated debate,

P, JAMES BAKER

Phoenix Therapeutics,16 Verbena Close, Kimberley Gardens,Canning Town, London El 6 4NU, UK

ReferencesBAKER, P, J, & WALSH, T , (1993) Smoking and cogni-

tive perfonnance: data derived from the temporalorder discrimination paradigm. Unpublished manu-script, London Guildhall University, June 1993,

LEVIN, E, D , (1992) Nicotinic systems and cognitivefunction, Psychopharmacology, 108, pp, 417—431,

WEST, R, (1993) Beneficial effects of nicotine: fact orfiction? Addiction, 88, pp, 589-590,