Impact of Use of Alcohol and Illicit Drugs by AIDS Patients on Adherence to Antiretroviral Therapy in Bahia,Brazil

Abstract

Use of alcohol and illicit drugs is a common finding among HIV-infected individuals, but there are many open questions about its impact on adherence to antiretroviral therapy and virological outcomes. Our study aimed to evaluate the impact of the use of alcohol and illicit drugs on the adherence to antiretroviral therapy (ART) among patients starting ART in Salvador, Brazil. We followed up 144 AIDS patients initiating ART for a 6-month period. At baseline, they were interviewed about demographics, behavior, and use of illicit drugs and alcohol. All of them had HIV-1 RNA plasma viral load and CD4⁺/CD8⁺ cells count measured before starting therapy. After 60 days of treatment they were asked to answer a new questionnaire on adherence to ART. All patients were monitored during the following months, and new CD4⁺ cell count/HIV-1 RNA plasma viral load determinations were performed after 6 months of therapy. Optimal adherence to therapy was defined by self-reported questionnaire, by 95% use of prescribed drug doses, and by using plasma HIV-1 RNA viral load as a biological marker. A total of 61 (42.4%) patients reported alcohol use, 7 (4.9%) used illicit drugs, and 17 (11.8%) used both alcohol and illicit drugs. Being in a steady relationship was protective to nonadherence (95% CI: 0.18–0.84). Missing more than two medical visits was also associated with a 68% higher likelihood of nonadherence (95% CI: 0.10–1.02). After logistic regression we detected a higher risk of nonadherence for patients declaring use of alcohol plus illicit drugs (odds ratio=6.0; 95% CI: 1.78–20.28) or high-intensity use of alcohol (odds ratio=3.29; 95% CI: 1.83–5.92). AIDS patients using alcohol and/or illicit drugs are socially vulnerable, and need specific and flexible programs, combining mental health care, harm reduction strategies, and assisted drug therapy to maximize the chances of successful use of ART.

Introduction

T he introduction of highly active antiretroviral therapy (HAART) significantly reduced the morbidity and mortality of AIDS patients.¹ However, despite all the advances achieved in treatment simplification, adherence to antiretroviral therapy (ART) remains an important factor in treatment success rate.² Earlier studies estimated that the optimal adherence rate should be 90–95% of all prescribed doses.³ The evolution of drug potency, and the onset of new drugs/new classes of drugs, with different pharmacokinetic profiles, significantly changed this paradigm. Currently, the usual HAART regimens are more tolerant of variations in adherence, allowing more forgiveness for missing doses.⁴ Currently, it is estimated that viral suppression can be achieved with adherence levels varying from 54% to 100%, depending on the drug regimen used.⁵

In some studies intravenous drug use (IVDU) was associated with lower levels of adherence, but it was not confirmed by others.^6,7 Some available evidence also suggests that life-style is more important in predicting adherence to HAART than IVDU.^8,9 These questions can significantly interfere with the decision to start ART for such patients, due to concerns about their ability to correctly use the treatment, which usually retards the initiation of therapy.⁹ However, use of illicit drugs is not a barrier to a successful therapy, and proper management of such high-risk patients can improve their chances of viral suppression.⁷

Alcohol use is a common finding among HIV-positive patients worldwide. According to the World Health Organization (WHO), on average, 53% of patients admitted to having used alcohol in the previous month, and 8% fulfilled the criteria to be considered heavy drinkers.¹⁰ The impact of alcohol intake on HAART adherence has been controversial. It is important to note that most studies agree with the use of alcohol as an important predictive factor to nonadherence, and some of them detected a close association between lower adherence rates and frequency/intensity of alcohol intake.¹¹ Other evidence suggests that the intensity of alcohol intake is important in predicting nonadherence to ART.¹¹ A recent systematic review of 60 published articles (47 in HIV patients) about the effects of alcohol use on adherence to the therapy of chronic diseases found that six of seven studies of the highest quality on HIV patients showed a significant effect, with sizes ranging from 1.43 to 3.6.¹²

In Brazil, 52% of individuals older than 18 years declare to use alcohol. It is more frequent for men (65%) than for women (41%). Among drinkers, 60% of men and 33% of women had used more than five doses in their last drinking experience.¹³ Alcohol use was also associated with a higher vulnerability to HIV infection and to nonadherence to antiretroviral (ARV) therapy.^14,15

Due to these complex interactions between ART adherence, use of psychoactive drugs, and alcohol, it is important to know the frequency and characteristics of use of such substances among HIV-positive patients, and to estimate its impact on ART adherence. A better knowledge of these facts can help us to understand the relationships between them and the other factors involved. This work was intended to evaluate the relationship between alcohol intake or use of illicit drugs and adherence to therapy among a cohort of Brazilian patients initiating HAART.

Materials and Methods

We enrolled consecutive patients initiating ART between March 2006 and March 2008 who were attending the main referral center for AIDS care in Salvador, Brazil. The center cares for approximately 5,000 patients. All patients initiating therapy in the study period were invited to enter the protocol, once the following inclusion criteria were fulfilled: no previous ART, age equal to or higher than 18 years, and providing a signed informed consent. The exclusion criteria were mental illness or inability to attend frequent medical visits.

Following entry in the study, patients were asked to answer a standardized questionnaire, adapted from the CEBRID (Brazilian Center for the Study of Psychoactive Drugs) with two distinct sections: the first one was directed to sociodemographic characteristics while the second part was focused on alcohol and illicit drug use.¹⁶ In addition, a third section was included to assess the adherence levels, based on self-report during routine medical visits. We used a previously validated AIDS Clinical Trial Groups (ACGT) questionnaire and pill count (every 30 days, during medication delivery) performed by a trained pharmacist.¹⁷ Optimal adherence was defined as correct use of pills in a minimum of 95% of prescribed doses.

The drinking pattern was defined according to the standards of a Brazilian National Surveillance on the Pattern of Alcohol Intake, as follows.¹³ Frequency of alcohol use: frequent drinker: daily use of alcohol or 1–4 times/week; occasional drinker: 1–3 times per month; and sober: less than one use of alcohol per year.

The amount of alcohol intake was calculated by number of doses: one dose was defined as any drink intake containing 10–12 g of alcohol. The drinking intensity was calculated according to the number of doses ingested per drinking session: sober: no alcohol use; mild drinking: 1–4 drinking episodes in a month, regardless of the number of doses; heavy drinking: at least weekly drinking, more than five doses/session; and binge drinking: excessive use of alcohol in a short period of time; more than five doses qualified for being drunk.

A viral load measurement was performed at baseline and after 6 months of starting ART. A result of less than 50 HIV-1 RNA copies/ml of plasma was considered a biological marker of proper adherence to ART. Viral load determinations were performed by the b-DNA method (Siemens VERSANT b DNA HIV–1 RNA 3.0 Assay) according to the manufacturer's recommendations.

Descriptive statistics were generated, including means, standard deviations, and medians for quantitative data, and frequencies for categorical variables. Prevalence was expressed in percentages, and 95% confidence intervals were estimated based on binomial distributions. Bivariate associations between the primary outcomes (HIV-1 RNA<50 copies/ml of plasma, adherence rates) and exposures of interest, including drinking patterns, use of illicit drugs, and sociodemographics variables, were assessed using the chi-square or Fisher's exact tests. A multivariate analysis was performed using a logistic regression model (stepwise method, with ɛ=0.05) to control for potential confounding variables. All variables of interest that showed a significant association at bivariate analysis were included at first, and those not reaching statistical significance were excluded from the model after each of the following steps. Statistical calculations were performed using STATA (STATA Corp LP, College Station, TX). The protocol was reviewed and approved by the Research Ethics Committee from Bahia State's Secretary of Health.

Results

A total of 144 patients (59% males) were included in the study, which represents approximately 50% of the total number of individuals starting HAART in the study period. The median age was 37 years (range 19–77 years). Most (52.8%) were single (40.3%) or divorced (12.5%), were black/racially mixed (61,.8%), and had less than 8 years of formal education (72.3%). Table 1 summarizes the main sociodemographics characteristics of patients included in the study, according to the use of alcohol, illicit drugs, or both.

Table 1.

Selected Patient Characteristics and Association with Drug and Alcohol Use

	Participants n=144		Alcohol use			Illicit drugs use			Álcohol and drug use
Characterístics	n	%	n	%	OR (95% CI)	n	%	OR (95% CI)	n	%	OR (95% CI)
Gender					1.12 (0.51–2.20)			0.5 (0.11–2.28)			3.68 (1.07–12.60)
Male	85	59.00	37	43.53		3	3.5		14	16.47
Female	59	41.00	24	40.68		4	6.8		3	5.08
Age (years)					0.85(0.44–1.66)			3.02 (0.61–15.08)			1.75 (0.63–4.87)
<37	67	46.5	27	40.30		5	7.5		10	14.93
≥37	77	53.5	34	44.16		2	2.6		7	9.09
Race					0.17 (0.59–2.32)			0.82 (0.17–3.80)			1.15 (0.40–3.32)
Black/mixed	89	61.8	39	43.82		4	4.49		11	12.36
Other	55	38.2	22	40.00		3	5.5		6	10.91
Marital status					1.14 (0.59–2.23)			0.43 (0.08–2.21)			1.3 (0.47–3.58)
Married	68	47.2	30	44.1		2	2.9		9	13.2
Single	76	52.8	31	40.8		5	6.6		8	10.5
Education					1.24 (0.58–2.65)			0.48 (0.47–2.17)			0.64 (0.22–1.87)
Up to 8 years	105	72.9	46	43.81		4	3.8		11	10.48
<8 years	39	27.1	15	38.46		3	7.7		6	15.38
Employment					1.35 (0.67–2.68)			1.39 (0.30–6.47)			0.52 (0.16–1.68)
Yes	51	35.4	24	47.06		3	5.9		4	7.84
No	93	64.6	37	39.78		4	4.3		13	13.98
Individual income					1.04 (0.52–2.11)			0.65 (0.14–3.02)			4.26 (1.03–17.60)
≤1 minimum wage	96	66.7	41	42.7		4	4.2		15	15.6
>1 minimum wage	48	33.3	20	41.7		3	6.3		2	4.2
Family support					0.85 (0.41–1.61)			1.58 (0.30–8.37)			8.69 (0.31–2.44)
Yes	89	61.8	36	40.4		5	5.6		10	11.2
No	55	38.2	25	45.5		2	3.6		7	12.7
Living conditions					2.39 (0.99–5.74)			0.33 (0.07–1.45)			0.24 (0.09–0.64)
Regular	114	79.2	53	46.5		4	3.5		9	7.9
Homeless	30	20.8	8	26.7		3	10.0		8	26.7
Sexual behavior					0.57 (0.27–1.21)			2.06 (0.25–17.07)			0.57 (0.19–1.65)
Heterosexual	108	75.00	42	38.89		6	5.6		11	10.19
Homo/bisexual	36	25.00	19	52.78		1	2.8		6	16.67
STD					0.54 (0.27–1.06)			2.46 (0.48–12.58)			8.64 (2.33–32.06)
Yes	74	51.4	26	35.14		5	6.8		15	20.27
No	70	48.6	35	50.00		2	2.9		2	2.86

STD, sexually transmitted diseases.

Sixty-one (42.4%) reported alcohol use (AU), 7 (4.9%) were illicit drug users (IDU, two of them used intravenous drugs), and 17 (11.8%) were both alcohol and noninjectable illicit drug users (AIDU). The most used alcoholic beverage was beer (71.8%), followed by distilled drinks (19.2%) and wine (9.0%). The more frequently used illicit drugs were marijuana (three patients), marijuana and crack (two patients), and other drugs (two patients). Among those reporting use of alcohol and IDU, marijuana was used by all of them; nine also used cocaine (six patients used cocaine and crack) and one used crack.

Of note, we observed a significantly higher proportion of males (82.4%) among those patients that declared using both alcohol and illicit drugs, in comparison with those who used only alcohol (60.7%) or ID (42.3%, p=0.03). Familial support was higher among the IDU (71.4%) than among alcohol users (59%) or users of both (58.8%, p=0.03). In addition, significantly more AIDU (35.3%, p<0.001) were homeless, in comparison with groups of AU (8.3%) or IDU (28.6%). A history of previous sexually transmitted diseases (STD) was also more frequent among AIDU (88%) than ID (71.4%) or AU (51.4%). Self-reported consistent condom use (in all sexual intercourse) was significantly lower among AIDU (23.3%) and IDU (14.3%) than among AU (47.5%, p=0.02).

Optimal adherence was reported by 84 (58.3%) patients. Nevertheless, viral suppression (VL<50 copies HIV-1 RNA/ml of plasma) after 6 months was achieved by 92 (63.9%) patients. Adherence to ART was significantly associated with attending all scheduled medical visits. Among 105 (72.9%) patients who attended all visits, 70 (83.3%) reported 100% adherence, versus only eight (9.3%) of those who failed to attend one to three visits, and six (7.2%) who did not show up for more than three visits (p=0.004). Appropriate use of ART, accordingly to medical instructions (p=0.001), and not having drug holidays (p<0.001) were also significantly associated with optimal adherence (Table 2). In addition, having a steady relationship was protective against nonadherence to therapy: among patients who declared that they were engaged in a steady relationship (N=68) the nonadherence rate was 30.8%, but this proportion was significantly higher (51%) for those who declared not having a steady partner (OR=0.39, 95% CI: 0.17–0.84, p=0.017).

Table 2.

Self-Reported, Therapy-Related Patient's Behavior and Its Association with Adherence

	Sample		Optimal adherence
Behavior	n	%	n	%	OR (95% CI)
Missing medical visits in the previous 6 months					0.28 (0.13–0.58)
Yes	39	27.1	14	35.9
No	105	72.9	70	66.7
Use medications according to prescription					3.17 (1.60–6.27)
Yes	81	56.3	57	70.4
No	63	43.8	27	42.9
Drug holidays					0.48 (0.17–1.35)
Yes	26	18.1	5	19.2
No	118	81.9	39	33.1
Able to correctly identify which drugs were prescribed					1.60 (0.17–1.36)
Yes	36	25.0	24	28.6
No	108	75.0	60	71.4

Drinking frequency and intensity were significantly associated with self-reported adherence to ART, and to adequate viral suppression (p<0.0001 for both parameters/variables), in a bivariate analysis. On the other hand, binge drinking was significantly associated with lower self-reported adherence (p=0.005), but not with viral suppression (p=0.59). Table 3 displays these findings. However, in a multivariate analysis, the only factors significantly associated with nonadherence to therapy were the intensity of alcohol intake (OR=3.19, 95% CI: 1.85–5.79, p<0.001) and use of alcohol plus ID (OR=5.99, 95% CI: 1.77–20.31, p=0.004).

Table 3.

Pattern of Alcohol Use and Its Relationship to Adherence to Antiretroviral Therapy

	Patients n=144		Optimal adherence
	n	%	n	%	OR (95% CI)
Drinking frequency					0.11 (0.03–0.31)
Frequent	28	19.40	4	14.30
Occasional	50	34.70	31	62.00
Intensity of use					0.12 (0.03–0.34)
Heavy drinking	27	18.80	4	14.80
Moderate use	51	35.40	31	60.80
Binge drinking					2.33 (0.80–6.86)
No	58	40.30	29	50.00
Yes	20	13.90	6	30.00

Discussion

We detected a prevalence of 54.2% of alcohol use among Brazilian patients initiating HAART in Salvador, Brazil. In addition, 16.7% also used any ID, in association with or not associated with alcohol. These rates are consistent with that found for the general population (52%) in previous studies.¹³

Alcohol use is a common finding among HIV-positive patients worldwide.^10,18
–20 It is important to note that most studies agree with the use of alcohol and illicit drugs as important predictive factors for nonadherence, and some of them detected an important association between lower adherence rates and frequency/intensity of alcohol intake.¹¹ Some other data suggest the intensity of alcohol intake is more important in predicting nonadherence to ART than frequency of use.¹⁸ In our study, the intensity of alcohol use and the association of alcohol intake and use of illicit drugs were strong predictors of nonadherence. However, use of only alcohol or illicit drugs was not associated with nonadherence to ART (Table 4).

Table 4.

Impact of Alcohol and Drug Use on Adherence to Antiretroviral Therapy and on Suppression of HIV-1 RNA Plasma Viremia

	Optimal adherence			Viral supression (viral load <50 copies/ml)
	n	%	OR (95% CI)	n	%	OR (95% CI)
Alcohol
Yes	29	34.50	0.46 (0.23–0.91)	39	42.40	1.01 (0.50–2.00)
No	55	65.50		53	57.60
Illicit drugs
Yes	3	3.60	0.51 (0.11–2.36)	2	2.20	0.20 (0.04–0.97)
No	81	96.40		90	97.80
Alcohol+illicit drugs
Yes	6	7.10	0.34 (0.12–0.95)	7	7.60	0.34 (0.13–0.95)
No	78	92.90		85	92.40

Male patients had a higher risk of being AIDU, and curiously, patients using ID were more likely to benefit from their family's support than AIDU or AU. In addition, more AIDU were homeless than people in the other categories. These facts suggest that relatives and friends usually consider ID use as a disease and provide a better social support to ID patients, but are less tolerant of alcohol use. This discrepancy can explain the different attitudes in the way they relate to AU (rejection) and IDU (support).

The proportion achieving optimal (95%) adherence (58.3%) was quite low, but the proportion of patients (63.9%) achieving viremia suppression after 6 months of therapy was slightly higher than this. These rates are lower than those observed in recent clinical trials involving drug-naive patients, which can usually reach 80–90% of efficacy at 6 months. However, it is compatible with the success rates observed in older clinical trials with a similar population of patients.

Adherence was also significantly associated with attending all scheduled regular medical visits and with proper use of ARV drugs (no drug holidays). One potential limitation of our study was the self-report measurement of adherence to ARV therapy, but this method has been largely used as a reliable way to measure this variable, including recent Brazilian studies.^21,22

Taken together, these findings suggest that the use of alcohol or even illicit drugs is not a major barrier to adherence to antiretroviral therapy, and patients reporting use of these drugs can benefit from ART in a manner similar to those individuals not reporting this type of behavior. However, the association of alcohol and illicit drugs use can significantly decrease the success rate of ART, due to its negative impact on adherence, and on viral suppression. Identifying patients with such a pattern of drug use can provide a chance to implement preventive measures in an attempt to minimize the risks of treatment failure due to nonadherence (Table 5). The presence of social support, expressed by being married/having a steady relationship, or having family ties was protective against nonadherence. This indicates that initiatives focusing on strengthening social support can help to improve the adherence rates and treatment outcomes for patients using alcohol and illicit drugs.

Table 5.

Multivariate Logistic Analysis of Factors Associated with Nonadherence to Antiretroviral Therapy in Salvador, Brazil

Risk fator	Z	Odds ratio	95.0% Confidence interval
Intensity of alcohol use	3.97	3.29	1.83–5.92
Use of alcohol and multiple substances	2.88	5.99	1.78–20.28
Being in a steady relationship	−2.38	0.39	0.18–0.84
Missing two or more medical visits	−1.92	0.34	0.09–1.02

The management and control of addiction to alcohol and illicit drugs are challenges for health care providers. The association of these problems with the need for treating HIV-1 infection is a hard task, and requires a multidisciplinary action, in order to reduce its negative impact on the treatment of chronic diseases, such as AIDS. Identifying the risk factors for nonadherence to therapy and identifying the exact role that alcohol and illicit drugs can play in therapy failure are basic requirements to design a proper strategy to manage patients affected by these problems.

Footnotes

Acknowledgment

This work was supported by a grant from Brazilian National AIDS Program, Ministry of Health.

Author Disclosure Statement

No competing financial interests exist.

References

Wong

, Chiasson

, Simonds

, Heffess

, Loo

. Antiretroviral therapy and declining AIDS mortality in New York City. J Urban Health, 2000; 77:492–500.

Moore

, Jeanne

, Gebo

, Lucas

. An improvement in virologic response to highly active antiretroviral therapy in clinical practice from1996 through 2002. Acquir Immune Defic Syndr, 2005; 39,2:195–198.

Paterson

et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med, 2000; 133:21–30.

Maggiolo

, Ravasi

, Ripamonti

, Gregis

, Quinzani

, Arici

, Airoldi

, Suter

. Similar adherence rates favor different virologic outcomes for patients treated with nonnucleoside analogues or protease inhibitors. CID, 2005; 40:158–163.

Bansberg

. Less than 95% adherence to nonnucleoside reverse transcriptase inhibitor therapy can lead to viral suppression. Clin Infect Dis, 2006; 43,7:939–941.

Ickovics

, Meisler

. Adherence in AIDS clinical trials: A framework for clinical research and care. J Clin Epidemiol, 1997; 50:385–391.

Esser

, Haberl

, Mulcahy

, Gölz

, Lazzarin

, Teofilo

, Vera

, Körber

, Staszewski

. Efficacy, adherence and tolerability of once daily tenofovir DF-containing antiretroviral therapy in former injecting drug users with HIV-1 receiving opiate treatment: Results of a 48-week open-label study. Eur J Med Res, 2011; 16,10:427–436.

Paterson

, Swindells

, Mohr

, Vergis

, Squier

et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Int Med, 2000; 133:21–30.

Organização Pan-Americana da Saúde. Guia para atenção e manejo integral de usuários de drogas vivendo com HIV/AIDS na América Latina e Caribe. 1. OPAS: Washington, DC, 2006.

10.

Abbey

, Zawackit , Buck

, Testa

, Parks

, Norris

et al. How does alcohol contribute to sexual assault? Explanation from laboratory and survey data. Alcohol Clin Exp Res, 2002; 26:575–581.

11.

Braithwaite

, McGinnis

, Conigliaro

, Maisto

, Crystal

, Day

et al. A temporal and dose–response association between alcohol consumption and medication adherence among veterans in care. Alcohol Clin Exp Res, 2005; 29:1190.

12.

Grodensky

, Golin

, Ochtera

, Turner

. Systematic review: Effect of alcohol intake on adherence to outpatient medication regimens for chronic diseases. J Stud Alcohol Drugs, 2012; 73,6:899–910.

13.

Laranjeiras

. Levantamento Nacional sobre os padrões de consumo de álcool na população brasileira. Secretaria Nacional Antidrogas, 2007.

14.

Silva

, Ximenes

, Miranda Filho

, Arraes

, Mendes

, Melo

, Fernandes

. Risk-factors for non-adherence to antiretroviral therapy. Rev Inst Med Trop Sao Paulo, 2009; 51,3:135–139.

15.

Bonolo Pde

, Machado

, César

, Ceccato

, Guimarães

. Vulnerability and non-adherence to antiretroviral therapy among HIV patients, Minas Gerais State, Brazil. Cad Saude Pub, 2008; 24,11:2603–2613.

16.

I Levantamento domiciliar sobre o uso de drogas psicotrópicas no Brasil: Estudo envolvendo as 107 maiores cidades do país: 2001. Carlini

E. A.

et al. São Paulo: CEBRID-Centro Brasileiro sobre Drogas Psicotrópicas: UNIFESP- Universidade Federal de São Paulo, 2002.

17.

Chesney

, Ickovics

, Chambers

, Gifford

, Neidig

, Zwickl

et al. Self- reported adherence to anti-retroviral medication among participants in HIV clinical trials: The AACTG adherence instruments. Patient Care Committee of the Adult AIDS Clinical Trials Group (AACTG). AIDS Care, 2012; 12:255–256.

18.

Parsons

, Rosof

, Mustanski

. Patient related factors predicting HIV medication adherence among men and women with alcohol problems. J Health Psychol, 2007; 12,2:357–370.

19.

Kenya

, Chida

, Jones

, Alvarez

, Symes

, Kobetz

. Weekending in PLWH: Alcohol use and ART adherence: A pilot study. AIDS Behav, 2013; 17:61–67.

20.

Fatch

, Bellows

, Bagenda

, Mulogo

, Weiser

, Hahn

. Alcohol consumption as a barrier to prior HIV testing in a population-based study in rural Uganda. AIDS Behav, 2012[Epub ahead of print]10.1007/s10461-012-0282-2.

21.

Santarem Ernesto

, Muller Banzato Pinto de Lemos

, Huehara

, Moreno Morcillo

, Dos Santos Vilela

, Nolasco da Silva

. Usefulness of pharmacy dispensing records in the evaluation of adherence to antiretroviral therapy in Brazilian children and adolescents. Braz J Infect Dis, 2012; 16,4:315–320.

22.

Sampaio-Sa

, Page-Shafer

, Bangsberg

, Evans

, Dourado Mde

, Teixeira

, Netto

, Brites

. 100% adherence study: Educational workshops vs. video sessions to improve adherence among ART-naïve patients in Salvador, Brazil. AIDS Behav, 2008; 12,4 Suppl:S54–62.