Clinical Preventive Services in Substance Abuse and Mental Health Update: From Science to Services

XIII. Appendix A: Literature Search Methods and Results

This Appendix supplements the information presented in the Methods section of this monograph. Details of the advanced searches conducted and their results, key words or search terms and methods used, and notes on selected search findings are presented here. In addition to the PubMed literature review for publications from 1964 through 2002, selected additional references were included in this report, as published between July 20, 2002, and October 27, 2003.

Keywords for Searches in PubMed for 1998–2002
The primary database used was PubMed. The following advanced searches were conducted:

Preventive health services OR preventive medicine OR preventive psychiatry OR primary prevention AND mental disorders NOT specific topics listed in items 2–9 below
Mass screening and mental disorders NOT in topics 3–9 below
Health education OR health promotion OR patient education AND mental disorders NOT topics 2, or 4–9
Home care services or home nursing AND mental disorders
Self-care and mental disorders (Note: there was no way to separately search on health risk appraisal in PubMed.)
Prenatal care OR perinatal care AND mental disorders
Disease management AND managed care AND mental disorders
Case management AND mental disorders
Psychoeducational (any reference where this term was used in title, abstract, or text; there is no MeSH term on this topic)

The literature search initially focused on the identification of pertinent and well conducted randomized controlled trials (RCTs). This was done to conduct the initial evaluation of interventions that deserve consideration for widespread implementation by health care systems. The search included the RCTs, literature reviews, and meta-analyses integrating data from multiple trials. Once this search was accomplished, additional literature on the selected topics was explored.

Table 3: Tabular Summary of Initial PubMed Search for 1998–2002

Topic	Initial Download	Selected to Pull Abstracts	Abstracts Pulled	Total RCTs^*	Pertinent RCTs¹
Preventive Health Services²	258	41	10	0	0
Mass Screening³	1,041	132	77	3	0
Health Education⁴	240	37	10	1	1
Home Care Services	314	36	20	5	3
Self-Care	426	105	17	6	5
Prenatal/Perinatal Care	139	56	26	4	4
Disease Management	482	99	68	15	14
Case Management	373	154	59	22	2
Psychoeducational	163	80	54	20	20
TOTALS	3,436	740	341	76	49

1. RCT = Randomized Controlled Trial
2. Excluding topics 2–9
3. Excluding topics 3–9
4. Excluding topic 2, and topics 4–9; limited to MeSH terms rather than all fields

Note: Since the search terms used for this initial search captured all the studies included relative to the six SAMHSA 2000 monograph topics, the numbers of abstracts and RCTs presented in this table, and the table immediately following, reflect only those not included in the SAMHSA 2000 search.

Table 4: Notes on Randomized Controlled Trials in Advanced Searches

Topic	Randomized Controlled Trials Rejected as Not Pertinent	Randomized Controlled Trials Considered Pertinent¹
Preventive Health Services²	(No RCTs)	The lack of RCTs appeared to be an artifact of the literaturesearch procedure.
Mass Screening³	Two were misclassified— they were not RCTs. The third was too poorly conducted to be of practical value (Schriger et al., 2001)	(None) It is important to note that the USPSTF guidelines on screening for depression are based on non-RCT studies of the screening procedures plus RCT studies of the efficacy of treatment of depression in persons detected by the screening procedures.
Health Education⁴	(None rejected)	Perry et al., 1999 (Perry, Tarrier, Morriss, McCarthy, & Limb, 1999)—successful RCT on educational program for patients with bipolar disease to reduce the frequency of manic relapse
Home Care Services	Two were purely therapeutic, with no preventive content	Two studies—Armstrong, Fraser, Dadds, & Morris, 1999, and Lagerberg, 2000—were preventive interventions to families with children considered at high risk because of social deprivation or “environmental factors.” Both showed positive results. Largerberg was a literature review. The third study—Gitlin, Corcoran, Winter, Boyce, & Hauck, 2001—was outreach to caregivers of patients with dementia, also showing positive results.
Self Care	One study (Pouwer & Snoek, 2001) dealt with diabetes, depression, and gender and appeared to be severely flawed	Three were meta-analyses or literature reviews—one each dealing with “adult problem behaviors,” dementia, and depression. The two RCTs addressed “chronically mentally ill outpatients” and anxiety attacks. All five are considered worthy of a closer look. The one on dementia is included to see if the intervention is for the patient or the caregiver.
Prenatal/Perinatal	(None Rejected)	Two RCTs dealt with depression, and one each with alcohol and drugs.
Disease Management	(None Rejected)	Nine of the 15 dealt with the cost-effectiveness of various approaches to treatment of depression. While therapeutic instead of preventive, these relate to the guideline to screen for depression. Of the other six, three dealt with alcohol, and one each with tobacco and depression.
Case management	20 of the 22 were purely therapeutic, with no preventive components	Of the two pertinent studies, one (Azrin & Teichner, 1998) was an instructional program to improve medication compliance for “chronically mentally ill” outpatients and the other (Buckwalter et al., 1999) dealt with a nursing intervention to decrease depression in caregivers of persons with dementia.
Psychoeducational	(None Rejected)	Most of the 20 studies in this group are therapeutic and lessen the progression of an illness or improve self-efficacy by the patient. Three have been pulled as “anchor” studies: Misri et al., 2000 (Misri, Kostaras, Fox, & Kostaras, 2000) deals with partner support in the treatment of postpartum depression. Von Korff et al., 1998 is a study of the treatment of depression Ostwald et al., 1999 (Ostwald, Hepburn, Caron, Burns, & Mantell, 1999) is an intervention for caregivers of patients with dementia

1. RCT = Randomized Controlled Trial
2. Excluding topics 2–9
3. Excluding topics 3–9
4. Excluding topic 2, and topics 4–9; limited to MeSH terms rather than all fields

Table 5: Searches From 1964 To 2002 To Collect Negative Studies

Topic	Papers Utilized as basis for Searches	Total References	Controlled Trials and Meta-Alalyses	Comment
Prenatal and Perinatal Home Visits	Olds Ramey Field	141 107 115	25 17 25	67 unduplicated abstracts (duplicates eliminated in Ramey and Field counts)
Tobacco Cessation	Marks	118	26
Short-Term MH Therapy	This category was deleted as a discrete category from this 2004 update, with the various interventions distributed to other categories not represented in the 2000 report.
HRA/Self-Care/Self Help	Kemper Vickery	251 109	54 48	102 unduplicated abstracts. The Vickery 48 exclude papers listed in Kemper search. One study by S. Moore (1980) showed no significant effects.
Presurgical Education	Devine Mumford	107 109	4 12	The Mumford 12 exclude papers listed in Devine Search.
Brief Education and Counseling To Reduce Alcohol Use	Fleming	172	Estimated-120	This is a high volume of studies, with the duration/length of the intervention and number of interventions per client highly variable and not well described in many papers.
TOTALS		1132	340

Table 6: Preliminary Analysis of SAMHSA 2000 Monograph Search for Negative Studies

	Topic	Pertinent Trials and Meta-Analyses (unduplicated)	Comment
D1	Prenatal and Perinatal Home Visits	67	Of the 67 unduplicated clinical trials and meta-analyses in this group of papers, 34 showed positive results, one negative results, and 32 were considered non-pertinent to this preliminary analysis—most because they were not home care or were not prenatal/perinatal visits. Of the 34 positive papers, Olds and/or Kitzman and Ramey authored 12. As a result, the number of actual clinical trials is less than the number of papers. In almost every instance, it seemed clear that the home care was part of a more comprehensive package of health and medical services—suggesting that simply adding a home care element to a straight clinical service is unlikely to be effective. Most of the studies could be classified along three general lines: General risk (economically and socially vulnerable groups) Drug/alcohol/tobacco users Children of mentally retarded mothers (A few of the studies dealt with infants with specific disorders or risk profiles) The benefits were mainly long-term social, psychological, and behavioral, rather than health care utilization. The babies were healthier and needed less long-term care. These studies were generally very well designed and showed strong positive benefits. The benefits, however, related to social dependency and issues other than offset of health care costs. While of obvious interest to health care systems serving Medicaid populations and health care providers serving economically vulnerable populations, these studies will be of relatively little interest to health care systems serving more well-to-do clientele.
D2	Tobacco Cessation	26	Of the 26 papers in this set, nine showed positive results, seven showed negative results, four were reviews, and six were not pertinent. The problem is one of getting an intervention that is intense enough to be effective—but not too costly—and then finding some way to extend the benefits beyond the end of the pregnancy.
D3	Short-Term MH Therapy	This category was deleted as a discrete category from this 2004 update, with the various interventions distributed to other categories not represented in the 2000 report.
D4	Self-Care/Self-Help	~100	More than 100 unduplicated papers are included in this self-help/self-care management data set; one negative study (randomized by family) by Moore (1980); and at least six positive studies, of which five used randomization. Most deal with education and training to help patients and family members do a better job of managing a medical chronic disease. The literature on managing chronic mental disorders is an entirely separate body of literature, with remarkably little overlap. Yet a third body of literature relates to the management of mental disorders as an aid to management of chronic medical conditions. This topic was merged into “Psychoeducation.”
D5	Presurgical Education	~16	Scanning the literature for relevant studies developed two nonrandomized clinical trials, one RCT, and two meta-analyses that were most relevant. Five were very positive. The large number of publications in this arena represents other and weaker study designs. This area has potential within the area of psychoeducation and the “activated” patient literature.
D6	Brief Education and Counseling To Reduce Alcohol Use	~120	Overall, two studies showed no effect of the intervention, while four were positive with strong results. This large data set includes a single meta-analysis (Poikolainen, 1999). This study concludes “for very brief interventions the change in alcohol consumption (6–12 month follow-up) was not significant among men or women.” For “extended brief interventions the reduction was statistically significant, but too small to be of clinical importance.” Most of the studies appeared to be more intensive.

Table 7: Topics from SAMHSA 2000 Monograph Reflected in This Monograph (SAMHSA 2004)

	2000 Topic Heading	2004 Topic Heading
D1	Prenatal and Perinatal Home Visits	High-Risk Pregnant Women and Children to Age 5
D2	Tobacco Cessation	Screening: Tobacco; Pregnant Women, Adolescents, and Adults
D3	Short-Term MH Therapy	Psychoeducation for Patients with Chronic Diseases
D4	HRA/Self-Care/Self-Help	Psychoeducation for Patients with Chronic Diseases
D5	Presurgical Education	Psychoeducation for Patients Scheduled for Surgical Procedures
D6	Brief Education and Counseling To Reduce Alcohol Use	Screening: Alcohol; Pregnant Women, Adolescents, and Adults

Table 5: Searches From 1964 To 2002 to Collect “Negative Studies” and More Recent Studies, Relative to the Interventions Recommended in the SAMHSA 2000 Report

These searches were conducted using an alternative PubMed search technique; that is, Keywords for Searches in PubMed for 1998–2002 The primary database used was PubMed. The following advanced searches were conducted:

Preventive health services OR preventive medicine OR preventive psychiatry OR primary prevention AND mental disorders NOT specific topics listed in items 2–9 below
Mass screening and mental disorders NOT in topics 3–9 below
Health education OR health promotion OR patient education AND mental disorders NOT topics 2, or 4–9
Home care services or home nursing AND mental disorders
Self-care and mental disorders (Note: there was no way to separately search on health risk appraisal in PubMed.)
Prenatal care OR perinatal care AND mental disorders
Disease management AND managed care AND mental disorders
Case management AND mental disorders
Psychoeducational (any reference wherethis term was used in title, abstract, or text; there is no MeSH term on this topic)

New topics added as a result of the broader literature review include screening of children and adolescents for evidence of behavioral disorder, screenings for illicit drug use, screening of adolescents for depression, and psychoeducation for persons with somatization.

Life Cycle Convention Used in This Report

After reviewing the literature and trying several alternative approaches, the following life cycle classification was used in this report:

Pregnant women
Pregnant women and mother-child dyads, birth to 5 years of age
Children 5–11 years of age
Adolescents 12–18 years of age
Adults 19 years of age and older (including seniors)

This approach compresses what otherwise would have been six or seven age groupings into five because the literature analysis showed that implementation guidelines were similar enough to warrant such compression. Pregnant women appear in two of the five groupings because of a discrete body of literature on the provision of preventive home visitation services to socially and economically vulnerable pregnant women, infants, and preschool children.

The adult population, previously divided into three age ranges (19–44 years; 45–64 years, and 65 years and older), was compressed into a single group after it became clear that the processes for screening, intervention, and follow-up basically were the same across all these age groups.

Although there are vast differences in risk profiles of children and adolescents, as they progress year by year from 5 to 18 years of age, the entire population in this lifestyle group tends to be neglected by the health care delivery system because of the group’s resiliency, generally good physical and mental health, and the ambiguity (which increases year by year into midadolescence) as to whether they or their parents are responsible for addressing risk-related and behavioral issues.

Format for Presentation of Guidelines

A common format is used for presentation of all basic interventions as follows:

Introduction and conclusions
Possible interventions
Review of literature: efficacy-evidence base for the guideline, including assessment of need, proof of efficacy, positive and negative studies, and studies addressing program implementation issues
Review of literature: effectiveness; program implementation issues; how to manage the intervention so that it succeeds in securing the desired benefits
Data to be tracked for surveillance, member selection, feasibility assessment, and program evaluation

TOC | Previous | Next