Behaviors, Vol. 11, pp. 389-397, Printed in the USA. All rights reserved.
Addictive
0306-4603/86 $3.00 + .OO Copyright o 1986 Pergamon Journals Ltd
1986
THE EFFECT OF WAITING FOR INPATIENT ALCOHOLISM TREATMENT AFTER DETOXIFICATION. AN EXPERIMENTAL COMPARISON BETWEEN INPATIENT TREATMENT AND ADVICE ONLY LASSE ERIKSEN University of Trondheim, ijstmarka Hospital, and Blue Cross Alcoholism Treatment Center, Trondheim, Norway Seventeen detoxified alcoholics were randomly assigned to a waiting list group or an inpatient treatment group. The waiting list group members were informed that they had to wait some time until there was a vacant place. By their discharge from the Detoxification Unit they were told to complete a self-report form every day with regards to drinking, working, sleeping home and use of disulfiram orally. Two weeks later they had an outpatient appointment in which the self-reports were collected and reviewed. This was repeated once for everybody so that they all had to wait 4 weeks. By discharge from further inpatient treatment, the inpatient treatment group was instructed to record the same four behaviors daily as the waiting list group did, and they, too, got an outpatient appointment with a 2 week interval. No significant differences between the groups were observed in the three main variables of drinking, working and sleeping home, but the waiting list group used significantly more disulfiram than the inpatient treatment group. Abstract-
Detoxified alcoholics from a detoxification unit at an alcoholism treatment center applying for further inpatient treatment usually have to wait some days or weeks on their own until there is a vacancy. Most of the staff of the center regard this waiting period as very disadvantageous for the alcoholics. Therefore, the effect of a waiting period of 4 weeks was studied by comparing a group on a 4-week waiting list with an inpatient treatment group for the first 4 weeks after discharge who received inpatient treatment immediately after detoxification. The behavioral view on alcohol problems, as opposed to the traditional medical model, presupposes that anyone included as a so called alcoholic will behave according to his contingencies of reinforcement (see Eriksen, Bates, & Gotestam, 1982). Hence, a waiting period with minimal treatment elements until inpatient treatment is offered, is not supposed to be automatically disadvantageous for the detoxified alcoholics, but, rather, present a possibility to make a change for a while. There are few relevant empirical reports approaching this problem in the literature. Pittman and Tate (1972) randomly assigned alcoholics to detoxification only (7-10 days) with no follow-up care and to detoxification plus 3 to 6 weeks of extensive inpatient treatment with extensive follow-up care. To assess the effects of the two treatment programs on subsequent social adjustment, seven dimensions of social functioning were utilized: drinking behavior, socioeconomic status, residential quarters, migration, general health, illegal behavior, and social stability. On all seven measures, both groups showed improvement at a 1 year follow-up in comparison to intake, with the direction of gains being greater for the experimental group. However, no significant differences
Requests for reprints should be sent to Lasse Eriksen, Department of Psychiatry and Behavioural Medicine, ijstmarka Hospital, P.O. Box 3008, N-7001, Norway. The present research was partially founded by the University of Trondheim. Knut Kvam, Sverre Saevareid and K. Gunnar Gotestam are gratefully acknowledged for assistance in different phases of the study. 389
390
LASSE ERIKSEN
between the experimental group (n = 172) and the control group (n = 78) were observed at 1 year follow-up. Stein, Newton and Bauman (1975) randomly assigned 58 alcoholics to detoxification only (for an average of 9 days) and to detoxification plus an intensive psychosocial inpatient treatment (for an average of 30 days). Posthospital adjustment was measured at five intervals over a 13 month period in the areas of social relationships, financial status, employment record, legal involvement, drinking behavior, use of community agencies, and readmissions to hospitals. In addition, measures were obtained on psychological change and counselling readiness. No significant difference was found between the two groups on any measure. The assessment measures in the present study were drinking behavior (abstinence, controlled drinking, or excessive drinking), being at work, sleeping home, and use of disulfiram orally. The being at work and sleeping home variables were included as measures of social functioning. Daily self-reports were chosen as the format of data collection because this was the only cheap and practical way to collect continuous information about the clients. In addition, daily reports from significant others (e.g., spouse/ cohabitant or parents) were collected when possible. The reliability and validity of self-reports in alcoholism research have often been questioned but a recent review and study by Polich (1982) shows that self-reports from alcoholics generally are valid. In addition, several studies have found that gathering data from clients’ significant others is an effective method for corroborating alcoholics’ self-reports of drinking behavior (Eriksen, Bjornstad, & Gotestam, 1984; Freedberg & Johnston, 1980; Maisto, Sobell, & Sobell, 1979; Miller, Crawford, & Taylor, 1979). The present study attempts to answer the question of the effect of waiting for inpapatient alcoholism treatment after detoxification. The hypothesis was that the outcome of the inpatient treatment group after discharge should be significantly different from the outcome of the waiting list group who, while waiting for inpatient treatment, receive only minimal outpatient treatment (daily self-reports and bi-weekly advice). In spite of this, it was hypothesized that the waiting list group in the interim would use more disulfiram than the inpatient treatment group after discharge from treatment. The reason for this second hypothesis was the experience of the alcoholism treatment center that almost none of the clients use disulfiram regularly more than a few days after discharge. It was further supposed that some of the clients waiting for inpatient treatment, however, would use disulfiram regularly to enhance the possibility of being sober when they got the offer of inpatient treatment. They were informed of the standard rule of the institution, which was admittance of sober alcoholics, so they would lose their opportunity for inpatient treatment if they were drunk the day they got the offer. METHOD
Clients and design Twenty-three voluntarily admitted clients of both sexes in a detoxification unit applying for short-term inpatient treatment at the treatment unit in the same building were asked to participate in the study. That includes every client during the 8 months the study was undertaken. One of the twenty-three clients refused to participate and another was imprisoned after a few days. Twenty-one thus gave their informed consent to participate in the study. The criteria for inclusion in the study were that the clients had finished their detoxification at the unit and were applying for inpatient treatment
Waiting
for inpatient
alcoholism
391
treatment
at the treatment unit, that they passed the usual criteria for admittance to this unit, that they lived within a specified geographic area (within a 3 hour drive from the center), and that they had a permanent address. The clients were randomly assigned to a waiting list group or an inpatient treatment group. Members in the inpatient group who ended their inpatient treatment before 3 weeks had passed, were excluded from the study because the treatment period was considered irregular or too short. Four clients were thus excluded. The final waiting list group had eight members, while the inpatient group had nine. The client characteristics of the two groups are shown in Table 1. The randomization was done in blocks of 10. The envelopes were pulled one by one, and each client was told the instructions accordingly. When drop-out occurred in the inpatient treatment group, a new envelope that indicated inpatient treatment was added to the rest of the envelopes to compensate for the drop-out. When the first female was assigned to one group, the next was placed in the other, as they were expected to be in minority. Instructions and treatment Waiting list group. The waiting list group members were told that they had to wait an unspecified time period before they could be admitted to the inpatient treatment unit, due to lack of capacity. They were told that they would be informed immediately when they could be admitted. In the meantime they were instructed to fill in a short daily selfTable 1.
Group
client characteristics
in numbers
or averages
Inpatient Treatment Group
Waiting List Group
N
9
8
Age
36.4 (28-52)’
32.6 (20-B)*
Sex
lF/SM
lF/7M
10.6 (7-14)*
9 (7-12)*
2 1 0 6
3 1 1 3
Education
(years)
Civil status Married/cohabitating Divorced/separated Widow/widower Single Previous
alcoholism
inpatient
status
8
6
Previous inpatient
alcoholism status last year
4
4
Number inpatient
of previous treatments
2.2 (O-6)’
1.6 (O-4)’
9 (l-14)*
7 (2-16)*
4
4
Years of problem Employment Inpatient
*range
alcoholism
drinking
at intake
period
(days)
47 (29-72).
392
LASSEERIKSEN
report concerning four behaviors. They were also given an appointment 2 weeks later. In this appointment the self-reports were reviewed, and the clients were informed that they had to keep on waiting. They were instructed to continue to complete daily selfreports in the same way as before, and they got a new appointment 2 weeks later. In the second appointment, after a total of 4 weeks of waiting, the self-reports were reviewed, and thereafter the clients were told that they now had the opportunity to join the inpatient treatment if they still wished. Two clients accepted inpatient treatment at once. Two additional clients accepted inpatient treatment, but wanted to postpone the treatment some months, because of summer vacation in one case, and inpatient treatment at a general hospital for the other. They were instructed to complete daily selfreports in the meanwhile. Four clients turned down the offer of inpatient treatment because they thought it was unnecessary. Three of these four clients were given outpatient treatment for a while. One of these applied for and was given inpatient treatment about 6 months later. The four clients in the waiting list group who did not accept inpatient treatment were instructed to fill in the daily self-reports for 3 months and were given appointments each month in the same time period. The two clients in the waiting list group who accepted inpatient treatment at once, the two clients who accepted but postponed their inpatient treatment, and the fifth client who received inpatient treatment 6 months later, were all instructed to complete daily self-reports, and were given appointments every month during the 3 month follow-up. Treatment group. The inpatient treatment group received the traditional short term abstinence oriented inpatient treatment which consisted of individual counselling, discussion groups, occupational training, recreational activities, physical training and lessons about alcohol and alcoholism. Their average inpatient stay was 47.0 days (range, 29-72) that is, about 7 weeks. The inpatient treatment group was instructed to complete the self-reports the first 3 months after their discharge. All clients in both groups were given pre-stamped and pre-addressed envelopes and self-report sheets. When they had filled in the self-report sheet, which covered 7 days, they should mail it. The members in the inpatient treatment group were given monthly appointments 3 months after discharge. In this appointment the self-reports were reviewed. Assessment and statistics Both groups were instructed to complete the daily self-report information sheet. This sheet required certain details of behavior for each day of the last week: how much the client had been drinking the last day (in categories of sobriety, controlled drinking, operationally defined as two drinks of liquor or less, or excessive drinking, that is, more than two drinks); whether the client had been at,work last day; whether the client had slept home last night; and whether the client used disulfiram (Antabus/Aversan) last day. They were shown how to complete the sheets. If a client had not filled in a self-report completely, this was completed during the appointment. If a client did not come to his appointment, he was contacted by telephone or mail for another appointment or for a visit at his/her home. Every client was followed-up 3 months either after inpatient detoxification (the six members of the waiting list group who did not accept inpatient treatment after 4 weeks waiting) or after inpatient treatment (all inpatient group members plus five waiting list members). If the client was living with his family or cohabitant, this significant other was asked to complete a parallel version of the daily self-report sheet and mail it every week. Four
Waiting for inpatient alcoholism treatment
393
significant others received a supply of self-report sheets and pre-stamped and preaddressed envelopes. The significant others recorded client behaviors in 27 weeks altogether. Applying the formula: r = 1 - actual deviations/possible deviations, the reliability for sober days was 0.95, working days 0.95, nights slept home 0.97, and disulfiram days 0.99. T-tests for independent groups were computed. Two-tailed tests were chosen for all variables except disulfiram days, because of the hypotheses and experimental design. RESULTS
The outcome of 28 waiting days for the waiting list group compared with the first 28 days after discharge for the inpatient treatment group is shown in Table 2. The only significant difference between the results of the two groups was on disulfiram days (t = 2.37, p < .05) when inpatient days were excluded (the waiting list group taking most disulfiram). Only two members in each group had any working days. In both groups two members were admitted to institutions because of alcohol problems. In the waiting list group four members consumed alcohol the first day after discharge from the detoxification unit, while five members of the inpatient treatment group consumed alcohol the first day after discharge. The outcome of the first 28 days after discharge for the five members of the waiting list group who also had inpatient treatment are shown separately in Table 2 (in fact these patients have a kind of reversal design, with one month waiting, a treatment period, and then 3 month follow-up). No significant differences appear between the results of this group and the waiting list group while waiting, nor the inpatient group after discharge. The same conclusion can be drawn when comparing 2 and 3 month follow-up of the inpatient group and the waiting list group after inpatient treatment (n = 5), and also the part of the waiting list group who waited at least 3 months before inpatient treatment (three members) or rejected this inpatient treatment (three members, in total n = 6). This is also shown in Table 2. (Three alcoholics are thus members of both the 3 month waiting list group and the waiting list group after inpatient treatment.) Controlled drinking days (defined as a maximum of two drinks of 40% alcohol or an equivalent amount of alcohol) were rare. The waiting list group reported 1 day altogether, the inpatient group 4 days and the waiting list group after inpatient treatment 4 days in the first 28 days. The low rate of controlled drinking days was expected because of the special population of the study from the detoxification unit, and because no controlled drinking training was implemented for any group member. In fact, the treatment goal of the center was complete abstinence, and every taste of alcohol was considered as a relapse. DISCUSSION
In the studied population of detoxified alcoholics, it does not seem disadvantageous to wait 4 weeks for further inpatient treatment under the present specified conditions (daily self-reports and bi-weekly advice). In fact, there was no significant difference while waiting between the waiting list group and the inpatient treatment group directly after discharge in all variables, except for disulfiram taking days. It can be argued, according to Kraemer (1981, p. 311), that with fewer than 10 subjects in each group, odds favor finding nonsignificant results even when the impact of the treatment is quite large. In the present study, the differences between the averages in the groups were mostly small, and, in addition, the ranges were considerable. In al-
19.8
7.9
0.25
6.25
Nights slept home”
Oral disulfiram” taking days
Admissions at institutions
Days spent at institutions 3.22
0.89
0
*Observations missing for one client ‘For one of the patients, the reports from the wife were used
“Inpatient days excluded
17.7
3.7
5.66
0.56
0
6.78
0.56
0.2
18.6’
8.4
Days on sick leave”
3.3
21.2
4.0
1.6
Working days”
-
14.0*
Days 51-84
-
9.6
6.8
Sober days just after discharge
15.6
Days 29-56
Inpatient Treatment Group (n = 9)+
11.6
17.8
17.8
Sober days”
Variables
Days l-28
WLG (n = 8)
5.20
0.40
1.4
10.8
6.2
5.8
1.4
12.6
Days 1-28
6.80
0.60
1.0
16.0
6.8
14.8
Days 29-56
6.20
0.40
0.6
18.6
-
6.0
-
18.2
Days 57-84
Waiting List Group After Inpatient Treatment (n = 5)
3.67
0.17
7.0
22.2
11.2
2.2
9.0
19.5
Days l-28
5.83
0.33
2.5
20.7
7.0
17.2
Days 29-56
5.61
0.33
0
20.2*
6.8
15.8*
Days 51-84
Three Months Waiting List Group (n = 6)
Mean outcome per group (Waiting List Group, Inpatient Treatment Group, Waiting List Group After Inpatient Treatment, Three Months Waiting List Group) and 28-day periods of follow-up
Days l-28
Table 2.
R
E
Waiting for inpatient alcoholism treatment
395
most every dependent variable extreme values were never obtained by only one mem-
ber; for example, when the inpatient treatment group (n = 9) had an average of 9.6 sober days just after discharge (in the first 4-week period) compared to the waiting list group after inpatient treatment (n = 5) with an average of 1.4 sober days just after discharge, the individual results show that in the first group five members drank within 24 hours and three members did not drink at all within the first 28 days, while in the other group three members drank within 24 hours and the other two after 3 and 4 days. It can further be argued that the waiting list clients were informed that they had to be sober when they got their inpatient treatment offer to be sure to get their admission. In addition, the waiting list clients did not know how long they had to wait. They were told they would be offered admission as soon as possible. These two conditions were, however, standard rules of the treatment center, and were not constructed for the study. Nevertheless, only two clients in the waiting list group accepted inpatient treatment immediately upon the end of the 4 weeks of waiting. Two other clients accepted inpatient treatment but wanted it postponed for 3 months. One client who was a former inpatient at an institution for old chronic alcoholics, was readmitted in his waiting period to the same institution, and was satisfied with that. Three of the eight clients in the waiting list group (37.5%) managed so well in the waiting period that they did not need inpatient treatment, but preferred outpatient treatment in agreement with their families. One of these clients was, however, admitted to inpatient treatment 6 months after initial detoxification. The other two clients were never admitted to inpatient treatment for 3 years afterwards and are reported to be managing well. In addition, the drop-out rate of inpatient treatment is noteworthy: four of thirteen clients (31 Ore)ended their inpatient treatment before 3 weeks had passed, thus indicating that this treatment was not suitable for them. The application of self-reports always raises the question of their trustworthiness. In spite of the good reports on reliability and validity of self-reports lately, one should always scrutinize the data in light of the methodology used. In the present study, there was a high degree of agreement between the reports of four significant others and the respective clients, but the other clients’ self-reports have not been corroborated. There is, however, no reason to suspect any of the clients of cheating on results. Some of the information was also occasionally checked with other informants and was never discordant. A change in the inpatient treatment program was implemented in the middle of the study period. The change consisted of a more planned and activity-oriented treatment program. The standard inpatient treatment period was changed from 4 weeks to 7 weeks. Three of the nine clients in the inpatient treatment group ended their inpatient treatment before the change of the treatment program, and one in the waiting list group. The number of drop-outs from inpatient treatment (with less than 3 weeks treatment) was equal in the two periods of the study (two in each period). A slight difference between the two groups with regards to age is apparent in Table 1. The waiting list group was about 4 years younger on the average than the inpatient treatment group. Four clients in the waiting list group were younger than the youngest in the inpatient treatment group (28 years old). It was not, however, a significant difference. The evaluation of the outcome in the present study compared to others is difficult since no other study of this kind has been found reported in the literature. Outcomes of some other kinds of studies may, however, have relevance to the present study. There are few reported outcome studies about the effects of detoxification only.
396
LASSE
ERIKSEN
Annis and Smart (1978) reported a follow-up of 522 alcoholics who were admitted to detoxification for the first time. Half a year follow-up revealed that 245 (46.9%) had been rearrested and 271 (52.0%) had been readmitted at least once. Annis and Liban (1978) reported a 3-month follow-up of 70 detoxified alcoholics. Half of them were (not randomly) assigned to a halfway house. Two alcoholics from the halfway house had been rearrested, and 17 had been readmitted for detoxification. Thirteen of thirtyfive matched controls had been rearrested and nine had been readmitted for detoxification. The total number of documented drunken episodes did not differ in the two groups. Pittman and Tate (1972) and Stein, Newton and Bauman (1975) found no significant differences in l-year follow-up between randomized groups of alcoholics who received detoxification only versus detoxification plus 3 to 6 weeks of extensive inpatient treatment. In an uncontrolled study Smart (1978) compared recovery rates for 174 alcoholics sent from a detoxification center to several different types of facilities for rehabilitating detoxified clients. Only about 35% were known actually to arrive, despite receiving their most preferred referral. Treatment and detoxification readmission data were gathered for all residents for a period of 12 months prior to and following their discharge from the detoxification center. A minority improved with regards to recidivism rates. Most had short lengths of stay in treatment, and they did not complete their treatment. The improvement rates did not differ, however, between those who arrived and those who did not arrive, or for those who completed the treatment and those who did not. No differences in improvement rates appeared for halfway houses, hospital or nonresidential programs. Inpatient treatment did not improve treatment outcome, but there are some signs that frequent outpatient treatment leads to improvement. In addition Smart et al. (1977) also showed that the type of referral did not relate to improvement. Two uncontrolled studies with 100 and 26 alcoholics respectively (Pattison, Coe, & Rhodes, 1969; Ritson 1968) found no significant difference in effectiveness between inpatient and outpatient alcoholism treatment programs assessed after 1 and several years follow-up respectively. Armor, Polich and Stambul(l978) reported in their multicenter (44) study with 2339 male clients followed-up 6 months after admission (and 597 male clients from six centers followed 18 months after admission) that there were no strong and consistent differences between the broad treatment categories (hospital, intermediate, outpatient). In a controlled study, Edwards and Guthrie (1966, 1967) randomly assigned 40 detoxified gamma alcoholics to either inpatient or outpatient treatment. Both groups received social counselling, family consultation, psychotherapy, AA-participation and prescriptions for calcium carbamide. No significant differences were reported between the groups at 6 (1966) or 12 months (1967). Edwards et al. (1977) (Orford & Edwards, 1977; Orford, Oppenheimer, & Edwards, 1976) reported a now classical study in which they randomly assigned 100 married alcoholic males to one of two treatment conditions after a thorough screening with tests and interviews, either an intensive several months outpatient treatment with or without inpatient treatment, or an “advice” group that received only single detailed and lengthy contact at the clinic, following which a social worker paid a monthly visit to the wife. In the “advice” session, the alcoholic was told that he should abstain, but that he himself would have to take responsibility for bringing about the change. No significant differences were found between the outcome for the two groups at either l- or 2-year follow-uns.
Waiting
for inpatient
alcoholism
treatment
397
The results from the present study are of considerable clinical importance. In spite of most of the staff’s predictions, the waiting list group on the average managed well during the 4-week waiting period, under the specified conditions (daily self-reports, and biweekly advice). It is further suggested that an obligatory waiting time period, two or four weeks, for example, under the same conditions would be very useful as a “motivation test’ or challenge” or a screening procedure to assign clients to the most appropriate treatment conditions (outpatient treatment, short-term intensive inpatient treatment, or longer, less intensive inpatient treatment). More studies exploring the results of the present study are warranted, especially with a larger number of clients, and with more homogenic groups than the sample of the present study. REFERENCES Annis, H.M., & Liban, C.B. (1978). A follow-up study of male halfway-house residents and matched nonresident controls. Journal of Studies on Alcohol, 40, 63-69. Annis, H.M., &Smart, R.G. (1978). Arrests, readmissions and treatment following release from detoxification centers. Journal of Studies on Alcohol, 39, 1276-1283. Armor, D.J., Polich, J.M., & Stambul, H.B. (1978). Alcoholism and treatment. New York: Wiley. Edwards, G., & Guthrie, S. (1966). A comparison of inpatient and outpatient treatment of alcohol dependence. Lancet, 1, 467-468. Edwards, G., & Guthrie, S. (1967). A controlled trial of inpatient and outpatient treatment of alcohol dependency. Lancer, 1, 555-559. Edwards, G., Orford, J., Egert, S., Guthrie, S., Hawker, A., Hensman, C., Mitcheson, M., Oppenheimer, E., & Taylor, C. (1977). Alcoholism: A controlled trial of “treatment” and “advice.” Journal of Studies on Alcohol, 38, 10041031. Eriksen, L., Bates, S., & Gotestam, K.G. (1982). Behavioral treatment of alcoholism: A review. Journal of Psychiatric Treatment and Evaluation, 4, 25-31. Eriksen, L., Bjbrnstad, S., & Gotestam, K.G. (1984). A lottery procedure to obtain alcoholics’ self-reports after discharge. Journal of Behavior Therapy and Experimental Psychiatry, 15, 147-151. Freedbere. E.J.. & Johnston. W.E. (1980) Validity and reliability of alcoholics’ self-reoorts of use of alcohol submitted before and after treatment. Psychological Reports, 46, 999-1005. _ Kraemer, H.C. (1981). Coping strategies in psychiatric clinical research. Journal of Consulting and Clinical Psychology, 49, 309-3 19. Maisto, S.A., Sobell, L.C., & Sobell, M.B. (1979). Comparison of alcoholics’ self-reports of drinking behavior with reports of collateral informants. Journal of Consulting and Clinical Psychology, 47, 106-l 12. Miller, W.R., Crawford, V.L., & Taylor, CA. (1979). Significant others as corroborative sources for problem drinkers. Addictive Behaviors, 4, 67-70. Orford, J., & Edwards, 0. (1977). Alcoholism: A comparison of treatment and advice, with a study of the influence of marriage. Oxford: Oxford University Press. Orford, J., Oppenheimer, E., & Edwards, G. (1976). Abstinence or control: The outcome for excessive drinkers two years after consultation. Behaviour Research and Therapy, 14, 409-418. Pattison, E.M., Coe, R., & Rhodes, R.J. (1969). Evaluation of alcoholism treatment: A comparison of three facilities. Archives of General Psychiatry, 20, 478-488. Pittman, D.J., & Tate, R.L. (1972). A comparison of two treatment programs for alcoholics. International Journal of Social Psychiatry, 18, 183-193. Polich, J.M. (1982). The validity of self-reports in alcoholism research. Addictive Behaviors, 7, 123-132. Ritson, B. (1968). The prognosis of alcohol addicts treated by a specialized unit. British Journal of Psychiatry, 114, 1019-1029. Smart, R.G. (1978). A comparison of recidivism rates for alcoholic detox residents referred to treatment facilities. Drug and Alcohol Dependence, 3, 218-220. Smart, R.G., Finley, J., & Funston, R. (1977). The effectiveness of post-detoxication referrals effects on later detoxication admissions, drunkenness and criminality. Drug andAlcohol Dependence, 2, 149-155. Stein, L.I., Newton, J.R., & Bauman, R.S. (1975). Duration of hospitalization for alcoholism. Archives of General Psychiatry, 32, 247-252.
