mg. daily to ensure maximum therapeutic response.
A trial period of one week of therapy is considered adequate
to determine the therapeutic effect of the drug. In the ab-
sence of a favorable response, therapy should be discon-
tinued.

Maintenance Dosage: When improvement is obtained, the
dosage should be promptly decreased to the minimum ef-
fective level which gives relief. This maintenance dosage
should not exceed 400 mg. per day because of possible cu-
mulative toxic effects. Satisfactory effect is often achieved
with as little as 100 to 200 mg. daily.

Dosage Regimen in Acute Gouty Arthritis: The treatment of
acute gouty arthritis calls for a somewhat different dosage
regimen. In this condition, best results are obtained from
an initial dose of 400 mg., followed by 100 mg. every four
hours until articular inflammation subsides, usually within
four days. This dosage should not be continued beyond a
period of one week.

Dosage Regimen in Acute Superficial Thrombophlebitis:
Best results have been obtained with a dosage of 600 mg.
per day in divided doses for the first 2 or 3 days, followed
by a reduction to 300 mg. daily. Usual duration of treatment
is 5 to 7 days and only rarely must therapy be extended be-
yond 10 days.

In severe forms of a variety of acute local inflammatory con-
ditions, satisfactory therapeutic results can usually be
achieved with a short 2 to 7 day course of therapy. Relief
of pain and resolution of swelling, redness and induration
is frequently achieved by these limited periods of treatment.
In general, 400 to 600 mg. daily in divided doses for the first
2 or 3 days is recommended. Subsequent doses of 300 mg.
daily are usually adequate.

In selecting the appropriate dosage in any specific case,
consideration should be given to the patient's weight, gen-
eral health, age, and any other factors that may influence
his response to the drug.

To minimize gastric upset, Tandearil should always be taken
immediately after food or with a full glass of milk.

Duration of treatment

The action of Tandearil is usually manifest by the third to
fourth day of treatment. It is not advisable, therefore, to
continue trial therapy with Tandearil beyond a week in the
absence of a favorable result.

In the treatment of transient conditions, such as painful
shoulder, medication should be discontinued a few days
following relief of symptoms; in the event of relapse, ther-
apy is resumed as needed to control symptoms. In the treat-
ment of more chronic disorders, medication may be con-
tinued at the minimal effective level required to maintain
relief and freedom from acute exacerbations. Medication

Electrolyte balance

Tandearil has a definite tendency to produce sodium reten-
tion and hence, edema, particularly in the older age group
of patients. Slight or moderate edema per se is not neces-
sarily a signal to cease therapy, since this complication is
usually self-limiting and invariably disappears on cessation
of treatment. Swelling of the ankles or face may often be
prevented or minimized by withholding salt from the diet,
or by reduction of dosage. Diuretics may be used if nec-
essary. In elderly patients, however, and in patients with
hypertension, cardiac defect or renal dysfunction, Tandearil
should be given with utmost caution, and the appearance
of clinical edema in such cases is an indication to discon-
tinue treatment.

Gastrointestinal tract

Nausea and vomiting are observed more frequently in pa-
tients who are forewarned about the possibility of such re-
actions, and this psychic factor should be kept in mind when
Tandearil is first prescribed. It is recommended that the
patient be instructed to take the medication immediately
after meals or with a glass of milk in order to minimize
gastric upset.

Although Tandearil has been found to be better tolerated
than Butazolidin, reactivation of latent peptic ulcer has oc-
curred. A detailed history should be obtained with special
emphasis on previous gastrointestinal disturbances. When
a history of peptic ulcer is elicited, Tandearil tablets should
be administered with utmost caution and only in treating
conditions of sufficient severity to warrant such usage. All
patients, regardless of their history, should be warned to
report immediately the occurrence of black or tarry stools.
Blood elements

Tandearil sometimes produces a moderate lowering of the
red cell count which is attributable to hemodilution rather
than to any direct effect on the erythrocytes or hemopoiesis.
Authenticated cases of agranulocytosis associated with the

drug have occurred. It is, therefore, imperative to guard
against this possibility by conducting routine periodic blood
counts. Any significant fall in the total white count or rela-
tive decrease in granulocytes should be regarded as a signal
for immediate cessation of therapy and institution of appro-
priate countermeasures. Thrombocytopenic purpura must
also be considered a possible side effect of Tandearil
therapy.

Allergic response

Drug rash occasionally occurs as a complication of Tan-
dearil therapy, but not infrequently it subsides when dosage
is reduced. The appearance of a mild rash, therefore, need
not necessarily be regarded as a signal for permanent dis-
continuance of treatment. When the rash is accompanied by
fever, a more serious degree of allergic response is indi-
cated, and the drug must usually be withheld. Purpuric rash
has been reported.

Rarely, a generalized type of allergic response occurs which
is similar to the serum sickness syndrome. This type of re-
action indicates a marked individual sensitivity to the drug
and medication must be permanently withdrawn.

Tandearil Geigy

(for complete Statement of Directions, please see inside)

initial: 300 to 600 mg. daily, in divided
doses, for 1 week. Usually unnecessary
to exceed 400 mg. daily.

maintenance: Minimum effective level,
often achieved with 100 to 200 mg.
daily, should not exceed 400 mg. daily.

400 mg. immediately, then 100 mg. every 4 hours until articular

inflammation subsides, usually within
4 days. Dosage should not
continue beyond 1 week.

initial: 600 mg. daily in divided
doses for 2 to 3 days.

maintenance: 300 mg. daily. Usual
duration of therapy is 5 to 7 days
(rarely 10 days).

initial: 400 to 600 mg. daily in divided
doses for 2 to 3 days.

maintenance: 300 mg. daily. Usual
duration of therapy is 2 to 7 days.

SOCIAL PROCESSES IN PHYSICIANS' ADOPTION OF A NEW DRUG

JAMES COLEMAN, CHICAGO, ILL..

HERBERT MENZEL, NEW YORK, N. Y.,

AND ELIHU Katz, Chicago, ILL.*

From the Bureau of Applied Social
Research, Columbia University

(Received for publication Sept. 26,
1958)

VER the past 20 years the practice of medicine has undergone profound changes, not the least of which has been the accelerated rate of change itself. Again and again, new diagnostic techniques, new laboratory tests, new drugs, new forms of anesthesia, new surgical procedures, and new principles of patient management make their appearance. Most of these innovations are minor steps which alter the medical scene only by gradual accretion, if at all; many, indeed, are short-lived or quickly superseded; a few have been milestones in medical developments. Whatever the ultimate significance of a new practice may be, its immediate fate, once it has been launched from the laboratory or research clinic, rests in the hands of the practicing physician in the field.

The pathways by which a successful innovation in medical practice spreads through the profession have seldom been systematically investigated. The present study is a contribution toward that end. It concerns the fate of a single innovation in four cities, a new variant in a well-established family of drugs. Such a case study can hardly claim to represent the course of all types of medical innovations under all circumstances, but it does illuminate some important paths

This is the first of two articles and may be identified as Publication No. A-273 of the Bureau of Applied Social Research, Columbia University. Preparation was facilitated by funds provided by a grant to the Bureau of Applied Social Research from the Eda K. Loeb Fund.

Philip Ennis, Marjorie Fiske, Rolf Meyersohn, and Joseph A. Precker participated in the planning of the study. Helmut Guttenberg and Sydney S. Spivack gave invaluable aid in the statistical analysis and in the collection of data, respectively.

*Coleman and Katz have transferred to the Department of Sociology, University of Chicago, since this study was conducted.

and processes through which medical innovations can make their way into the practice of physicians. One may add that drugs are peculiarly suitable as tracers of these paths, because they are physical objects with standardized names, their release dates are easily ascertainable, and pharmacists maintain exact prescription records.

Fig. 1.-Cumulative proportion of doctors introducing gammanym over a 16-month period (N = 125).

The data of this study stem from two sources. In four midwestern communities, interviews were conducted with 125 general practitioners, internists, and pediatricians, who generously contributed an average of 11⁄2 hours of their time. They constituted 85 per cent of all practitioners of their specialties in these cities. In addition, data on the prescriptions written by these 125 physicians were obtained from almost all the pharmacies in these communities.* The prescription data covered a period of 16 months beginning with the release date of a new drug which will here be called "gammanym." At this time two older drugs of the same general type were in widespread use. They had appeared some years earlier and are here designated as "alphanym" and "betanym." The three

*Ninety-one additional interviews were held with practitioners of other specialties, but their prescription record was not examined. The communities had altogether 356 physicians in active private practice. The sample was designed to include all the general practitioners, internists, and pediatricians, and a selected group of other specialists. The population of the four cities ranged from about 25,000 to just over 100,000.

medications belong to a well-established family of drugs which has widespread applicability in the hands of general practitioners and many specialists. By the end of the survey period, 16 months after its release, gammanym had become at least a part of the standard medication of most practicing physicians; 87 per cent of the general practitioners, internists, and pediatricians in the sample under study had introduced it into their practice. But this change had been neither immediate nor all-encompassing. The over-all rate of introduction of gammanym by these doctors is seen in Fig. 1. This cumulative curve indicates for each stated date the percentage of doctors who had already prescribed the drug up to that time. As can be seen, more and more doctors introduced the drug during this 16-month period until almost 90 per cent had used it; then the curve finally leveled off.

Fig. 2.-Overlapping use of alphanym, betanym, and gammanym during three time intervals.

To be sure, the rapid rise in the use of gammanym did not mean that its predecessors, alphanym and betanym, were dropped from use. The degree of overlapping use of two or three of the drugs by the same doctor is pictorially represented in Fig. 2, for three time intervals, representing the beginning, middle, and end of the period studied. Thus, for example, as late as 16 and 17 months after the release of gammanym,* only 22 per cent of the doctors were prescribing gammanym exclusively, while 15 per cent were prescribing both betanym and gammanym, 13 per cent were prescribing both alphanym and gammanym, and 20 per cent were writing prescriptions for all three drugs during these same 2 months (more accurately, during the 6 sampling days representing these 2 months). The introduction of gammanym meant for most doctors an addition to whatever drugs of this type they were already using, rather than a substitution. Only slowly, if at all, did some of the doctors stop using the older drugs.

DRUG ADOPTION AND INDIVIDUAL CHARACTERISTICS OF PHYSICIANS

This, then, describes the over-all course of the acceptance of gammanym in these four communities, from the time it first arrived on the drugstore shelves

*The prescription data were 3-day samples at intervals averaging 28 1⁄2 days, which are here termed "months," so that the 16-month survey period contains 17 such "months." Because of the nature of time sampling, the introduction dates recorded here may be somewhat later than the physicians' actual introduction of the drug.