wm and li report

Since Boston's Long Island Shelter first opened its doors in 1982, it has expanded both its bed capacity and its service approach. Begun with 100 beds at one location, it now provides almost 700 beds in two locations; initially meeting only the emergency needs of guests for food, shelter and medical care, it now offers a physical health clinic, a psychiatric nurse, substance abuse counseling and job and training programs. The shelter also has opened a home that provides transitional living for eight former guests.

Service delivery for most of Long Island's guests begins with an intake interview administered by a case manager. The intake interview identifies general service needs and particular concerns for each guest--a profile that can then inform subsequent decision-making by nurses, counselors and other shelter staff. For years, statistical analysis of these intake interview data has helped to describe the diverse population of shelter guests and identify their unmet needs.

By 1990, the shelter's growth had begun severely to tax the capacity of the existing system for collecting, processing and analyzing intake interview data. At the same time, shelter management began to detect changes in the service needs of shelter guests that suggested a need for a more systematic assessment. As a result, the shelter decided to sponsor a comprehensive survey of its guests at both locations: the Long Island Shelter itself and the newer Woods-Mullen Shelter that is administered by Long Island. This report presents the basic survey findings and draws some comparisons with the earlier studies.

METHODOLOGY

Guests at the Long Island and Woods-Mullen shelters were asked to participate in the interview study between July and November, 1990. Guests were identified for interviews using a stratified random sampling technique. Nightly bed lists provided the sampling frame for the study. On selected days and nights each week, guests whose bed numbers matched a randomly generated list of numbers were approached by a staff member or a research assistant and asked if they would consider participating in an interview. Each prospective respondent was offered a pair of socks or a pack of cigarettes in exchange for participating, and each signed a detailed consent form before the interview began. The acceptance rate was 80 percent.

Shelter staff helped to recruit guests for the interviews, which were conducted by specially trained research assistants from the university's graduate programs in clinical psychology and applied sociology; interviews were conducted in special rooms at both locations and required from 45 to 90 minutes.

Many guests at both shelter locations participated in a "contract bed program." Both special programs within the shelter and arrangements to facilitate activities outside of the shelter are operated as contract bed programs: a shelter-based stabilization program for alcoholics, a work experience program within the shelter, enrollment in courses at a local college or training institute, or simply regular employment outside of the shelter. Some contract beds also are allotted to individuals with physical or mental health problems who need special care. Contract bed participants are able to keep their bed assignment each night and to leave some belongings at the shelter during the day; participants in in-shelter programs are also able to remain in the shelter during the day.

Contract bed participants and other guests were sampled separately, in part to take advantage of the availability of many program participants for daytime interviews at the shelter, and to ensure proportional representation of both types of shelter users. The final sample was then weighted to represent exactly the proportions of guests at the two shelters and in each program at both shelters.

The interviews consisted of approximately 150 questions concerning respondents' residential history, satisfaction with the shelter, physical and mental health, substance abuse, employment and benefits, social supports and social background. Questions were drawn from two sources: the shelter's intake interview schedule and, in some areas, standard scales. The following scales were included in the interview schedule: the CES-D (a measure of depression); the PERI Scale of False Beliefs and Perceptions (psychotic thinking); the ISEL (a measure of social support); the xx program satisfaction scale; and selected questions from the ASI (the Addiction Severity Index). An extensive series of AIDS awareness questions was also included in the interview schedule. The Life Lines AIDS Prevention Project for the Homeless collaborated in developing these questions.

In addition to describing the distributions of most variables, this report identifies differences between guests staying at Long Island and Woods-Mullen. The relationship between numerous sets of variables also was investigated, and selected relationships are included in the report. All reported differences were statistically significant at the .05 level or better; in other words, there is less than one chance in twenty that any association between two variables that is mentioned in the report was due to chance, rather than to an actual association in the total population of shelter guests.

BACKGROUND

More than half of the shelter users were younger than 40; one-quarter were in their twenties. Nine percent of the shelter users were at least 60 years old. The median age was 35.

AGE

Percent

20s 26%*

30s 33

40s 16

50s 17

60s 9

101%

(152)

*One respondent was 19 years old; she is included in the "20s" category in this table.

Four in five shelter users were men. More than half were black, while one-quarter were white and about one in ten were Hispanic.

SEX

Percent

Male 79%

Female 21

100%

(155)

RACE

Percent

White 26%

Black 57

Hispanic 11

Asian,Pacific 1

Am. Indian,Alaskan 1

Other 4

100%

(154)

Prior schooling varied widely among the shelter users. About one-third had gone no further than grade school, another third had completed high school, and one third had had some college (although very few had completed college). About one in five reported some difficulty reading.

LAST GRADE OF SCHOOL COMPLETED

Percent

No school, less than 8 5%

Grade school 28

High school 34

GED 8

Some college 16

Vocational school 3

College 1

Graduate work 5

100%

(154)

HOW WELL CAN YOU READ?

Percent

Read well 81%

OK 15

Not at all 4

100%

(155)

READING DIFFICULTIES

Any trouble reading....

Yes

Road signs 15%

Newspaper 21

Each of the demographic characteristics reported was associated with some health problems and other characteristics. Men reported less social support than women--two-thirds of the women had at least a moderate level of social support available, compared to just under half of the men. Those who had not graduated from high school were more likely to have been incarcerated--three-quarters compared to just over half. The two education groups also differed in satisfaction with the shelter: two-thirds of those who had not completed high school were satisfied with the shelter, compared to just under half of the high school graduates. Self-reported problems with reading had one significant correlate: those who reported problems with reading were less likely to be looking for work (50%) than those who did not report problems with reading (28%). Veterans were much more likely than other men to report having been assaulted or robbed in the past year--two-thirds compared to about one-third.

Race

Variable Min. White

Contract Bed 39% 72%

Interested--AIDS ed.53% 30%

Age

Variable <=30 31+

Use Drugs 46% 17%

Health problem 30% 52%

Health insrnce 44% 63%

Victimized 24% 50%

Legal problem 39% 18%

AIDS interest 63% 40%

Good health 75% 55%

THE SHELTER EXPERIENCE

Most respondents used the shelter intensively: more than half reported staying at the shelter "most of the time" since their first night at the shelter and two-thirds had stayed there every night during the previous week.

FREQUENCY OF SHELTER USE

Since first stay, you've stayed at this shelter...

Percent

Most of the time 52%

Some of the time 38

Only rarely 10

100%

(154)

NUMBER OF NIGHTS STAYED HERE LAST WEEK

Percent

0-1 10%

2-3 6

4-5 13

6-7 72

101%

(154)

Almost half of the respondents had participated in a contract bed program, in which they received a regular bed and some other privileges in exchange for participating in a program. The specific program in which respondents had participated was not identified in many cases, but the most often mentioned programs were employment in or outside the shelter (13 percent) and health care (8 percent). About four percent had participated in an in-shelter program for alcoholics.

Program Enrolled in or Finished

Program Frequency Percent

Work Exp 7 4.7

Lit Prog 1 .5

Formal Sch 1 .5

Stablz/Holding (Alc) 5 3.8

Work outside 6 4.0

Emp Serv 6 4.1

Heal C 12 7.6

Other Prog 23 14.5

Program not known 28 18.1

Not in Program 65 42.1

------- -------

Total 155 100.0

More guests at the Woods-Mullen Shelter were participating in a contract bed program than at the Long Island Shelter--about two-thirds compared to one-third.

A series of questions were asked to identify respondents' level of satisfaction with the shelter. The responses indicated general satisfaction, with complaints about some particular shelter features.

Four in five shelter users were satisfied overall with the shelter in which they were interviewed; almost one-quarter were very satisfied and just under one in ten were very dissatisfied. When compared to other shelters, two-thirds thought their shelter was better and only one in ten found it to be worse. Nonetheless, when compared to their last regular housing, almost two-thirds said the shelter was worse--what is perhaps most striking is that more than one-quarter of the shelter users found the shelter to be at least somewhat better than their last regular housing.

OVERALL SATISFACTION WITH SHELTER

Percent

Very satisfied 23%

Satisfied 50

Dissatisfied 19

Very dissatisfied 9

101%

(149)

COMPARISON OF SHELTER TO OTHER SHELTERS/LAST REG. HOUSING

Other Shelters Last R. Housing

Much Better 34% 13%

Somewhat Better 31 16

About the Same 24 11

Somewhat Worse 6 25

Much Worse 5 36

100% 101%

(122) (143)

Satisfaction varied with specific shelter features. Services at the shelter were rated as excellent by one in five and as at least good by more almost three in five. About one-third thought the shelter's services were just fair, but only six percent rated services as poor.

Satisfaction was highest with shelter staff, with the shelter's physical conditions and safety and with health services at the shelter. At least seventy percent were satisfied in these areas. Just over half the shelter users were satisfied with the shelter's convenience to stores and shopping, with the help the shelter provided with getting a job, and with the other residents of the shelter. Satisfaction was lowest with respect to the amount of privacy at the shelter--just over four in ten were satisfied in this area and one-quarter were very dissatisfied.

RATING THE QUALITY OF SERVICES AT SHELTER

Percent

Excellent 22%

Good 37

Fair 35

Poor 6

100%

(154)

LEVEL OF SATISFACTION WITH SHELTER CHARACTERISTICS

V. Dis- V.

Sat. Sat. Sat. D.Sat. Tot.

Condition/state of repair 17% 63 14 6 100%

Amount of room/space 11% 65 18 6 100%

Furnishings 11% 56 25 9 101%

Staff 19% 61 12 8 100%

Amount of privacy 6% 37 32 25 100%

Security/safety 18% 53 18 12 101%

Other residents 8% 49 25 18 100%

Convenient for shopping 16% 43 25 17 101%

Health services 20% 49 21 10 100%

Help with getting a job 8% 49 27 16 100%

N=89-149

Levels of satisfaction differed between Long Island and Woods-Mullen with respect to several satisfaction dimensions: Long Island guests were more satisfied with the level of privacy at the shelter; Woods-Mullen guests were more satisfied with the shelter's condition and location.

SERVICES

Shelter users relied on a wide array of service agencies for meeting their needs. Four in ten had had contact with the Department of Public Welfare. Both Social Security and other shelters had been contacted by about one-quarter of the respondents in the preceding year. Between 10 and 15 percent of the respondents had had contact in the last year with the Boston Housing Authority, the St. Francis House day program and the Department of Social Services. Other agencies and organizations had been used less often.

Most respondents who had used another service agency or organization in the preceding year reported that they received the services they needed. Over 90 percent had received the services they needed from other shelters and the St. Francis House, while between 80 and 90 percent had received the services they needed from DPW. Among the agencies contacted by at least 10 percent of the respondents, satisfaction was lowest for two: the Boston Housing Authority and the Department of Social Services. Nonetheless, about two-thirds reported that they had received the services they needed from these agencies.

Long Island Shelter guests reported somewhat less contact with service agencies than did Woods-Mullen guests.

CONTACT WITH SERVICE AGENCIES IN PAST YEAR

Contact Rcvd. Serv. Needed

Dept. Pub. Welfare 41% 85%(62)

Social Sec. Admin. 25% 82%(38)

Other homeless shelters 23% 93%(36)

Boston Housing Authority 14% 68%(21)

St. Francis 13% 95%(18)

Dept. Soc. Services 11% 63%(16)

Vocational rehab 7% 66%(11)

Mass. Rehab. Comm. 7% 71%(10)

BCH Soc. Serv. 7% 85%(10)

Veteran's Admin. 6% 56%(9)

Dept. of Mental Health 5% 90%(8)

Other 5% 75%(8)

Salvation Army 3% 100%(4)

Bridge 3% 100%(7)

Agency for Elderly Af. 1% 100%(1)

Dept. of Immig. 1% 100%(1)

Traveler's Aid 1% 100%(2)

Comm. for Blind 0%

RESIDENTIAL EXPERIENCE AND PREFERENCES

About one in five shelter guests had stayed with friends and another one in five with family members since they had used the shelter. A comparable proportion had stayed in other shelters or halfway houses during this same period; just one in ten had stayed on the streets since first using the shelter.

OTHER PLACES STAYED SINCE FIRST STAY

Percent

Friends 18%

Family 20

Own Place 11

Streets 11

Oth Shel or Hlfway Hse 23

Institution 3

Other 5

Have only Stayed Here 9

100%

(154)

One-quarter of the respondents had been homeless for no more than four months; the median length of time homeless and the median length of time since first staying at the shelter was one year. More than one-third had been homeless for at least two years. Dorchester, Roxbury and the South End were the neighborhoods where most shelter users had most recently had a regular place. Just seventeen percent came from outside the Boston metropolitan area; two-thirds had most recently lived in the city of Boston itself.

LENGTH OF TIME HOMELESS

Percent

Under a Month 7%

1-4 Months 20

5-11 Months 17

1 Year-19 Months 18

2-3 Years 21

4 Years or More 16

99%

(139)

LAST REGULAR PLACE

Percent

South End 10%

Back Bay 4

South Boston 1

East Boston 3

Boston Proper 5

Roxbury 16

Dorchester 21

Jamaica Plain 2

Cambridge 1

Chelsea 1

Brookline 3

Beacon Hill 2

Roslindale 1

Mattapan 3

Malden 1

Revere 1

South Sub. Bost. 5

North Sub. Bost. 4

West Sub. Bost. 1

Western Mass. 3

Other N.E. States 1

Outside N.E. 12

Outside the U.S. 1

102%*

(143)

*Error due to rounding

In spite of their overall satisfaction with the shelter, the respondents were eager to move on. Almost two-thirds said they wanted to move out "right now," and all but five percent wanted to move out within at least one year. The overwhelming favorite residential preference was for an independent apartment, rather than a group living arrangement. Almost nine in ten respondents "definitely" wanted to get their own apartment rather than live with other folks from the shelter.

Two-thirds felt "distressed" about not having a place to stay, although fifteen percent said they felt "OK" about not having a place to stay.

PREFERRED LENGTH OF STAY AT SHELTER

Percent

Rest of my life 5%

Several more years 0

6 months-a year 15

Several wks.-a month 19

Like to move out now 61

100%

(149)

If given the choice, would you rather share a house w

other folks from the shelter, or get your own apartment?..

Percent

Definitely share house 2%

Probably share house 2

Don't really care, not sure 4

Probably get own apartment 7

Definitely get own apartment 86

101%

(154)

FEELINGS ABOUT NOT HAVING A REG. PLACE

Percent

Distressed 65%

So-So 20

O.K. 15

100%

(155)

SOCIAL SUPPORT

Although few respondents were married, most had some family connections. The majority of respondents had never been married and about one in five were divorced; only one in ten were married or living together. Nonetheless, almost two-thirds had relatives in the Boston area and two-thirds reported that they received a great deal or a fair amount of support from family members. Two in five reported children in the Boston area and a similar proportion had children elsewhere. Most of these children were in the care of a relative and about one in ten were in foster care.

Friends were an important source of social support for the shelter users. Most of the respondents, four in every five, had good friends in the Boston area and 62 percent received a great deal or a fair amount of personal support from friends.

Shelter staff also provided social support to many shelter users. Almost half of the respondents reported a supportive relationship with someone on the shelter's staff and two-thirds received a great deal or a fair amount of personal support from shelter staff. Overall, more than two-thirds reported having a special person they could count on in times of need.

MARITAL STATUS

Percent

Married 7%

Living Together 3

Separated 6

Divorced 23

Widowed 4

Single 58

100%

(154)

FAMILY AND FRIENDS IN THE AREA/ELSEWHERE

Boston Area Elsewhere

Have Children 42% 38

Any Relatives 64% 73

Have Good Friends 79% 57

N=105-145

CARE OF CHILDREN

Percent

Foster 12%

Relative 52

Adopted 2

On Own 9

Else 25

100%

(49)

SPECIAL SOURCES OF SUPPORT

Yes Not No Tot.

Sure

One Spec. Person 71% 5 23 99%

Someone on Shel Staff 47% 23 30 100%

N=155

GAUGING PERSONAL SUPPORT

Amount of Personal Support Received From......

Great Fair Not Too None Tot.

Deal Amount Much

Shelter Staff 23% 45 15 18 101%

Friends 28% 34 18 20 100%

Family 40% 17 16 27 100%

N=145-148

Social support was also measured with responses to a series of statements that comprise the widely used "ISEL" index of social support. Respondents could answer "definitely true," "probably true," "probably false," or "definitely false" in response to such statements as: "When I feel lonely, there are several people I can talk to."

The median response to the ISEL statements was 2.9, indicating that the respondents were most likely to feel that it was "probably true" that they had social supports. The ISEL distribution was somewhat clumped toward the low support end of the scale, but only one respondent received the lowest possible average score.

ISEL Social Support

Count Value One symbol = approx. 4.00 occurrences

1 1.00

35 2.00 *********

102 3.00 **************************

16 4.00 ****

....I.........I.........I.........I

0 40 80 120

Histogram frequency

Overall, guests staying at Woods-Mullen reported somewhat higher levels of available support than did those staying at Long Island.

PHYSICAL HEALTH ASSESSMENT

Physical health problems were common, although by no means universal. Almost half reported at least one current physical health problem and one-third had a medical prescription; more than half had seen a doctor in the preceding year and more than one-third had been hospitalized; almost one-third were at least somewhat dissatisfied with their physical ability to get things done; almost one-fifth rated their own physical health as worse than most. Of course, this suggests that a majority of shelter users had no physical health problems and were satisfied with their physical health. In any case, about one-third reported they had no health insurance.

PHYSICAL HEALTH COMPARED TO OTHERS

Percent

Better than most 43%

About the same 42

Worse than most 16

101%

(152)

SATISFACTION WITH PHYSICAL ABILITY

Percent

Completely satisfied 29%

Very satisfied 21

Somewhat satisfied 20

Somewhat dissatisfied 15

Very dissatisfied 9

Completely Dissatisfied 6

100%

(155)

HAVE ANY PHYSICAL HEALTH PROBLEMS NOW? 45%

SHOULD BE TAKING PRESCRIBED MEDS? 36%

HAVE ANY HEALTH INSURANCE? 36%

N = 154-155

HOSPITALIZED OR SEEN BY DOCTOR FOR PHYS.

HEALTH PROBLEM IN PAST YEAR

Percent

Hospital only 17%

Hospital and doctor 22

Doctor only 14

No 48

101%

(155)

The specific health problems ranged from heart and circulation problems (approximately five percent mentioned them) to bone problems and injuries (seven percent) and eye/ear/nose/throat problems (four percent).

HEALPROB Types of health problems

Freq Percent

PSYCHIATRIC 2 1.2 PANCREAS KIDNEY BLAD 1 .5

HEAD INJURIES 2 1.5 SEIZURE DISORDER 2 1.5

DIABETES ALLERGIES B 2 1.4 CHEMICAL DEPENDENCY 1 .4

OTHER INJURIES 2 1.2 EYE EAR THROAT 3 2.2

CIRCULATION 8 5.0 EYE EAR THROAT RECEN 3 1.7

HEART CIRCULATION 0 .3 AIDS 1 .5

HEART STOMACH 2 1.4 MUSCLE 3 1.7

HEART MUSCLE 1 .5 BACK 2 1.0

HEART RECENT OPERATI 1 .5 DIABETES 1 .5

STOMACH 1 .7 RECENT SURGERY 2 1.0

KIDNEY OR BLADDER 1 .5 OBESITY 1 .7

BONE 7 4.4 OBESITY PICKWICK SYN 1 .5

GAUT ARTHRITIS 1 .5 THYROID 2 1.5

ASTHMA 1 .5 DENTAL 3 2.0

ASTHMA EPILEPSY 3 1.7 STROKE 1 .5

ASTHMA BACK EYE EAR 1 .5 SLEEP DISORDER 1 .5

HIGH BLOOD PRESSURE 1 .5 NO HEALTH PROBLEM 86 55.2

HEALTH PROBLEM NOT ID'D 5 5.0 TOTAL 155 100.0

Prescribed medications were for the same range of problems, with seven percent also mentioning medication for psychiatric problems.

What is medication for?

Value Label Frequency Percent

SEIZURES 3 1.9

SEIZURES THYROID 1 .5

HEART 1 .5

NERVOUS CONDITION 4 2.3

NERVOUS CONDITION HE 2 1.4

DEPRESSION 5 3.0

SCHIZOPHRENIA 1 .5

BACK 7 4.5

HEART 1 .8

PREGNANCY 2 1.0

BLOOD PRESSURE 7 4.7

BLOOD PRESSURE KIDNE 1 .5

BLOOD PRESSURE GLUCO 1 .5

HYPERTENSION LIVER D 1 .5

MOLE 3 1.7

ASTHMA 2 1.2

KIDNEY TUMOR 2 1.4

BACK 1 .5

DIABETES 1 .5

PNEUMONIA 1 .5

SLEEP CONDITION 1 .5

LIMBS 2 1.6

NO MEDICATION 98 63.0

MEDICATION NOT ID'D 10 6.2

----- -------

Total 155 100.0

Poorer physical health was associated with several other variables. Those who reported physical health problems were more likely to be receiving financial benefits, but less likely to be employed and less likely to be looking for work. Those who were ill also knew less about AIDS.

Health Problem

Variable Yes No

Financial benefits 58% 23%

Looking for work 30% 58%

Employed 13% 39%

Know AIDS facts 39% 56%

MENTAL HEALTH ASSESSMENT

About one in five respondents had been hospitalized for a mental or nervous problem and one in ten were patients at a community mental health center; about one quarter had been either in a psychiatric hospital or a CMHC. The median number of times hospitalized for those who had been hospitalized was 2.0.

TREATMENT FOR MENTAL OR NERVOUS PROBLEM

Yes

Been Hospitalized for Mental or Nervous Problem? 21% (153)

Assigned to Community Mental Health Ctr in Boston? 9% (151)

NUMBER OF TIMES HOSPITALIZED FOR MENTAL OR NERVOUS PROBLEM

Percent

1 21%

2 31

3 19

5 10

6 8

20 12

101%

(25)

Respondents were asked numerous questions about their feelings. Three quesitons came from the lengthy "Mental Health Index"; they indicate the proportion of guests who felt unhappy and who had had suicidal thoughts.

Between 20 and 30 percent of the respondents had felt unhappy or depressed in the previous month, and 14 percent had thought of taking their own life at some time in the previous month. Between one-quarter and one-third had been generally happy in the previous month.

How happy/satisfied over past month

Percent

EXTREMELY HAPPY 11.5

VERY HAPPY/MOST TIME 12.3

GENRLLY SATISFIED 21.9

TIMES HAPPY/UNHAPPY 30.2

GENRLLY UNHAPPY 13.2

VERY DISSATISFIED 10.9

Missing

-------

100.0

(149)

So down nothing could cheer you up in past month

Percent

ALL THE TIME 6.2

MOST OF TIME 12.7

GOOD BIT OF TIME 10.9

SOME OF TIME 30.0

A LITTLE OF TIME 25.3

NONE OF TIME 14.9

-------

100.0 (149)

Think of taking your life in the past month?

Percent

YES, CONSTANTLY 2.0

YES, VERY OFTEN 3.3

YES, FAIRLY OFTEN .5

YES, A COUPLE OF TIM 5.0

YES, ONCE 2.5

NO, NEVER 86.6

-------

100.0 (149)

An index of depressed feelings was constructed from six questions selected form the widely used CES-D index. Response choices varied from 1 to 3, with 3 indicating depressed feelings "all of the time," 2 indicating depressed feelings "some of the time," and 1 indicating depressed feelings "never." The median depression score was 1.7, and the cases tended to clump just under the middle section.

CESD

Count Midpoint One symbol = approx. .50 occurrences

16 1.0 ********************************

0 1.1

10 1.2 ********************

14 1.3 ***************************

0 1.4

17 1.5 **********************************

1 1.6 **

25 1.7 *************************************************

21 1.8 ******************************************

0 1.9

17 2.0 **********************************

0 2.1

10 2.2 *********************

5 2.3 **********

0 2.4

6 2.5 ************

0 2.6

6 2.7 ***********

5 2.8 ***********

...I....+....I....+....I....+....I....+....I

0 5 10 15 20

Frequency

Guests at the Long Island Shelter reported more symptoms of depression than did guests at Woods-Mullen.

The PERI scale of False Beliefs and Perceptions (with yes/no response choices) indicated that relatively few guests harbored false beliefs or perceptions that might indicate psychosis or hallucinations for other reasons. The median number of false beliefs and perceptions was two of a possible thirteen.

PERI

Count Value One symbol = approx. 1.00 occurrence

43 .00 *******************************************

20 1.00 ********************

23 2.00 ***********************

23 3.00 ***********************

10 4.00 **********

9 5.00 *********

8 6.00 ********

4 7.00 ****

1 8.00 *

4 9.00 ****

0 10.00

0 11.00

1 12.00 *

3 13.00 ***

...I.........I.........I.........I.........I

0 10 20 30 40

Prior hospitalization for psychiatric problems was associated with more symptoms of depression, but a greater chance of having received financial benefits in the past month. Such individuals were also more worried about getting AIDS but knew less about prevention methods.

Prior Psych. Hospitalization

Variable Yes No

More depressed 64% 39%

Thought of suicide 25% 10%

Financial benefits 54% 34%

Worry about AIDS 86% 64%

Know AIDS facts 30% 53%

Those who reported more false beliefs and perceptions were also more likely to have had difficulty affording things--58% compared to 38% of those with few false beliefs and perceptions.

TRAUMA AND ABUSE

Traumatic events were remarkably common. Two in every five respondents had been assaulted or robbed in the past year and two-thirds had experienced a life-threatening experience at some time in their lives. Three in every five of these persons experienced nightmares about losing their life at least rarely (half had such nightmares at least sometimes).

PHYSICALLY ASSAULTED OR ROBBED IN PAST YEAR?

Percent

Assaulted 11%

Robbed 15

Both Assaulted and Robbed 17

No 58

101%

(153)

NIGHTMARES ABOUT LOSING LIFE

Freq. of Nightmares Related to Trauma in Past Yr...

Percent

Frequently 13%

Sometimes 23

Rarely 22

Never 42

100%

(112)*

*Question asked of the 67% who reported a life-threatening experience.

Many respondents had first been traumatized as children. One-third had been physically assaulted by an intimate as a child and two in five had run away from home or been in foster care.

CARE AS A CHILD

Percent

Foster Care 12%

Instit. Care 6

DSS 2

No Such Care 80

100%

(155)

CHILDHOOD EXPERIENCE

Yes

Phys. Assaulted by Pers. Close to you? 31% (153)

Ran Away from Home/Foster Care? 41% (153)

Any Childhood Trauma

Variable Yes No

False beliefs 69% 43%

Psych. hospitlztn 32% 13%

Drinking problem 32% 15%

Drug problem 35% 14%

Health problem 53% 34%

Legal problems 32% 15%

Victimized 50% 32%

Childhood trauma was related to a number of more recent problems: more false beliefs and perceptions, a much higher rate of pscyhiatric hospitalization, more evidence of drinking and drug problems, more physical health problems, more legal problems and more victimization while homeless.

SUBSTANCE ABUSE

Three-quarters of the sample drank at least occasionally, half had been told at some time that they had a drinking problem, and two in five had been treated for a drinking problem. One-third admitted to drinking at least several times a week at the time of the interview.

FREQUENCY OF ALCOHOL CONSUMPTION

Percent

Daily 20%

Several Times a Week 17

Once a Week 11

2 or 3 Times a Month 13

Less Than Once a Month 13

Never 26

100% (152)

PROBLEM DRINKING

Yes

Anyone Ever Said You Drink Too Much? 50%

Ever Been Treated for Drinking Prob? 39%

N=118-119

Indicators of drinking problems were associated with a variety of other characteristics and orientations. Drinkers were more likely to have had suicidal thoughts and to be depressed. Drinkers reported more difficulty affording things and were less likely to be receiving financial benefits. In addition, drinkers were more likely to have been in the custody of youth services as an adolescent.

Drinking Problem

Variable Yes* No**

More depressed 60% 40%

DYS involved 31% 12%

Difficult to afford 72% 45%

Financial benefits 18% 46%

*N=33-38

**N=112-6

Street drug use was less common. Half said they did not use any street drugs in the previous year; of those who admitted to some street drug use, half had been treated at some time for a drug problem.

FREQUENCY OF STREET DRUG USE IN PAST YEAR

Percent

Daily 14%

Several Times a Week 7

Once a Week 7

2 or 3 Times a Month 11

Less Than Once a Month 11

Never 51

101%

(151)

DRUG TREATMENT

Yes

Ever Been Treated for Drug Problem? 46%(88)

The correlates of drug abuse were more limited than were those for alcohol abuse. Those with a drug problem had more legal problems and greater difficulty affording things. Drinking problems were much more common among drug users than among others.

Drug Problem

Variable Yes* No**

Legal problems 46% 17%

Difficult to afford 67% 47%

Drinking problem 45% 17%

*N=150-153

Overall, substance abuse appeared to be more common at the Woods-Mullen Shelter than at Long Island.

AIDS

AIDS awareness was an area of particular concern in the survey. Shelter users were asked about their worry about getting AIDS, their beliefs about HIV infection, and about their exposure to and reactions to various shelter-based prevention activities. Their responses indicated the importance of this topic to many.

Almost half of the respondents worried "a lot" about getting AIDS, although over one-third worried not at all. One-third of the respondents believed that other shelter residents worry a lot about AIDS. Almost half said they knew someone with AIDS. In response to other questions (not in the table presented here), 43 percent of the respondents reported that they were afraid around people infected with HIV, but only ten percent thought that they'd probably get AIDS. Sixty-four percent wanted to learn more about AIDS.

Responses to another question also revealed the widespread concern about AIDS among shelter users: more than one-third had been tested for HIV infection and another third wanted to be tested.

WORRY ABOUT GETTING AIDS

A lot Some- None Tot.

what
How much do you worry about AIDS? 46% 17 37 100%

Other residents worry about AIDS.. 35% 39 26 100%

N=137-150

KNOW ANYONE WITH AIDS?

Yes 43%

Not Sure 9

No 49

101%

(150)

AIDS TESTING

Have Been Tested 38%

Like to Get Tested 36

Not Sure 8

Would Not Like Testing 17

99%

(151)

About nine in ten of the respondents knew the most basic facts about HIV transmission: it can be transmitted through sex and sharing needles and it can't be transmitted just through proximity or casual contact. However, half believed that giving blood could transmit HIV and between 15 and 25 percent mistakenly identified insects, toilet seats, kissing and food preparation as risk behaviors.

Responses to another series of statements concerning AIDS beliefs and knowledge also reflected almost universal awareness of the most basic facts but much confusion beyond these. At least nine in ten knew that preventive action can reduce the risk of infection, that using a condom helps, that risk increases with the number of sex partners and that AIDS does not only affect gay men. Slightly fewer respondents knew that only a blood test tells if you're infection, that a man can get AIDS from sex with a woman, that people don't die soon after infection and that a condom can still help even after infection.

There was more disagreement about a range of other AIDS beliefs. About two-thirds believed, correctly, that there's no cure for AIDS, that drug users can avoid AIDS and that infected people can still feel healthy.

AIDS KNOWLEDGE

Percent Who Believe the Following Can Give You AIDS....

Being Near Infected People 9%

Hugging an Infected Person 8

Being Bitten by an Insect 25

Having Sex w Infected 90

Sharing Needles w Drug Users 90

Giving Blood 50

Toilet Seat Used by Infected 16

Kissing an Infected Person on Mouth 29

Eating Food Prepared by Infected 20

N=142-146

BELIEFS ABOUT AIDS

Percent Agreeing with the Following Statements About AIDS...

There's No Cure for AIDS 65

Infected People Can Still Feel Healthy 66

AIDS Only Affects Gay Men 7

The More Sex Partners, More Risk of AIDS 92

No Way Drug Users Can Avoid AIDS 29

It Helps a Lot to Use a Condom 90

Only a Blood Test Tells if Infected 84

People Can Do a lot to Avoid Infection 96

Men Can't Get AIDS from Sex w Woman 15

People Die Soon After Infected 29

A Condom Doesn't Help After Infection 78

N=132-142

Overall, the median percent correct on the AIDS knowledge questions was 75. The responses were clumped toward the high end of the distribution, indicating many correct answers to the AID knowledge questions.

KNOWAID

Count Value One symbol = approx. .80 occurrences

1 30.00 **

7 40.00 ********

10 50.00 *************

13 60.00 ****************

28 70.00 ***********************************

32 80.00 ****************************************

36 90.00 *********************************************

17 100.00 *********************

I.........I.........I.........I.........I

0 8 16 24 32

Histogram frequency

Valid cases 143 Missing cases 12

There had been a substantial amount of AIDS prevention activity at the two shelters, but awareness of these activities was by no means universal. About half had read prevention information at the shelter and an equal proportion had noticed condoms available at the shelter. Just one in five had noticed bleach bottles available for drug users and only five percent had received a greeting card with a condom (and a prevention message)--a prevention activity sponsored by an AIDS prevention project that had occurred once at the shelter. Few had attended an AIDS awareness meeting at the shelter, but then few had heard of such meetings being held.

AIDS PREVENTION ACTIVITIES

Have Read Info. at Shelter 50%

Noticed Condoms Available 52

Received Card w Condom 5

Noticed Bleach Bottles 18

N=144-152

Attended AIDS Education meeting?

Attended 18%

Haven't Heard About One 69

Heard, Didn't Go 13

100%

(153)

About two-thirds of the respondents were comfortable with participating in the most straightforward prevention activities: picking up or requesting prevention supplies; speaking to a shelter staff member about AIDS; and attending a meeting about AIDS.

COMFORT WITH AIDS PREVENTION ACTIVITIES

Feel Comfortable......

Taking Condom from Basket 62%

Asking Male Staff for Condom 66

Asking Female Staff for Condom 65

Taking Bleach from Basket 46

Asking Staff Member for Bleach 63

Speak to Staff About AIDS 72

Would Attend AIDS Meeting 79

N=139-151

How effective were these activities? One in five respondents reported that the shelter's AIDS education efforts had affected their behavior or the behavior of other shelter users. Almost three-quarters believed that shelter staff should do more to educate shelter users about AIDS. And more than two-thirds felt that people with AIDS were treated more poorly by other shelter users.

EVALUATION OF AIDS PREVENTION

% Agree

People w AIDS are Treated Poorly by Others 69%

Staff Should Do More Educ. Guests re AIDS 73

AIDS Education Affected Own Behavior 21

AIDS Education Affected Others 23

N=145-152

LEGAL HISTORY

One-quarter of the respondents were currently on probation, parole, pending arraignment or involved in a legal action in some other way. Two-thirds had been in jail or prison at some time in the past and one in five had been under DYS supervision when they were younger.

CURRENT LEGAL STATUS

Percent

Probation 9%

Parole 7

Pending Arraign. 3

Other 6

None of the Above 76

101%

(153)

BEEN IN JAIL/PRISON?

Percent

Jail 44%

Prison 20

No 36

100%

(150)

INVOLVEMENT WITH DYS

Yes

Involved w DYS When Younger? 18% (152)

WORK AND BENEFITS

About half of the respondents had not had difficulty affording food, clothing or health care in the past year, while a third had experienced such difficulties at least pretty often.

FINANCES OVER PAST YEAR

Difficulty Affording Food, Clothing or Health Care

in Past Year

Percent

Never or Hardly Ever 28%

Once in a While 20

Sometimes 21

Pretty Often 13

Always or Almost Always 17

99%

(151)

Three quarters were not working, and two in five of the unemployed were looking for work. Almost all of the respondents had worked in the past and almost half of these had spent at least five years at one job. Almost one-third had earned more than $10 per hour in at least one previous job; the median maximum hourly wage was $8.50 (the mean maximum hourly wage was $10.73.

EMPLOYMENT STATUS

Percent

Full-Time 15%

Part-Time 11

Day Labor 1

Not Working 73

100%

(154)

UNEMPLOYED

Yes

Looking for Work 43%

Have Worked in Past 93

N=117

LONGEST PERIOD HELD A JOB

Percent

5 or More Years 44%

1-4 Years 42

6-11 Months 9

1-5 Months 3

Less than 1 Month 2

100%

(144)

HIGHEST WAGE EARNED PER HOUR

Percent

$3.00-5.50 13%

$6.00-8.00 33

$8.50-10.00 24

$10.75-13.75 14

$14.00-20.00 10

$20.00 + 6

100%

(137)

Guests at Woods-Mullen were more likely to be employed than were guests at Long Island.

BENEFITS

Two in five respondents were receiving financial benefits. Social Security Income was the most common benefit, with welfare and general relief a close second. Few persons received any other type of financial benefit. No more than one respondent in ten reported having a representative payee who received the respondent's benefits.

BENEFITS RECEIVED

Percent

Welfare 19%

Disability 4

Unemployment 2

SSI 39

V.A. Benefits 7

Gen. Relief 15

Food Stamps 7

Grants 2

Illegal 2

Disability plus SSI 4

99%

(54)

REPRESENTATIVE PAYEE

Percent

Yes 9%

No 88

Don't Know 3

100%

(146)

SERVICE NEEDS

The importance attached to receiving help varied markedly across the 13 services mentioned. By far the most importance was attached to getting help with finding a place to stay: four in five considered it extremely important. The next most important service needs were help with financial benefits and AIDS education--about three in five respondents considered help in these areas extremely important.

Somewhat less importance was attached to receiving more education, help with physical health problems, job training and help with finding a job. Two in five respondents rated these forms of help as extremely important, although one-quarter rated each of them as not at all important. About one-third rated help with drug problems and with literacy as extremely important, but almost half rated these forms of help as not at all important.

Help with making more friends and with psychological problems were rated as extremely important by only one-quarter and as not at all important by almost one-third. The forms of help that were rated as least important were ehlp with an alcohol problem and help with legal problems: one-quarter rated them as extremely important but almost three in five rated them as not at all important.

IMPORTANCE OF HELP WITH SERVICE NEEDS

Not Slight. Mod. Consid. Ex- Tot.

At All trem.

Finding a Place 6% 6 6 4 79 101%

Financial Benefits 12% 6 7 16 58 99%

AIDS Education 13% 10 10 10 56 99%

Education/GED 27% 9 6 12 47 101%

Finding a Job 28% 14 7 7 45 101%

Physical Health 29% 13 8 9 42 101%

Job Training 28% 12 8 13 40 101%

Drug Problems 45% 11 4 5 35 100%

Literacy 45% 7 9 9 30 100%

Psych. Problems 29% 21 11 13 27 101%

Alcohol Problems 58% 9 7 3 23 100%

Legal Problems 57% 5 5 5 28 100%

Making More Friends 30% 14 21 14 21 100%

N=88-154

Getting help with alcohol and drug problems were rated as important by more guests at Woods-Mullen than at Long Island.

CONCLUSIONS

Guests' sociodemographic characteristics varied markedly, but were similar in some respects to those found in earlier studies at Long Island and in research on other homeless populations. In particular, the median age of 35 and the 20 percent who were women replicated findings in many other studies. The proportion who were black or Hispanic, a figure that varies more between cities and regions, had risen markedly from earlier studies at Long Island, from forty to seventy percent, due in part to the opening of the Woods-Mullen Shelter.

Resources available to the shelter guests were, if anything, even more variable. The educational levels of grade school, high school and college were represented in equal proportions, and substantial proportions had low, medium and high levels of social support. Family and shelter staff were important particularly as sources of social support.

About equal proportions reported some and no physical health problems. Symptoms of depression were common, symptoms of distorted perceptions were uncommon; about one in five had been psychiatrically hospitalized, as reported in many other similar studies. Substance abuse was common.

Other areas for concern were the high rate of reported legal problems and criminal victimization, with even higher rates of reported abuse as a child.

Several guest characteristics were associated with a variety of other characteristics and problems: substance abusers, those who had been abused as children, and, for certain problems, younger guests were particularly at risk (older guests, however, were more likely to report physical health problems and criminal victimization). Those who had been hospitalized previously for psychiatric problems were currently more depressed, although they also were more likely to be receiving financial benefits.

AIDS was a major concern. Many worried about getting AIDS and knew someone with AIDS, many had been tested for AIDS and most wanted to learn more about AIDS. Knowledge of basic HIV transmission facts was widespread, but many were concerned needlessly about improbable transmission methods.

In most respects, the Long Island and Woods-Mullen shelters served the same population of homeless persons. However, there were some important differences, most of which could be explained by differences in the shelters' physical design and their locations--Woods-Mullen is located near the center of Boston, next to Boston City Hospital, while Long Island is on an island in Boston Harbor that can be reached only by car or bus.

Woods-Mullen had a higher rate of participation in the contract bed program for workers; related to this, more of them reported being employed. Woods-Mullen guests were more satisfied with the shelter's location and its physical condition (it had recently been refurbished), while Long Island guests were more satisfied with the shelter's level of privacy (beds were separated in clusters by partitions, while at Woods-Mullen they were arranged dormitory style). Woods-Mullen guests reported more contact with service agencies and somewhat higher levels of social support. Woods-Mullen guests reported more substance abuse but fewer symptoms of depression. Woods-Mullen guests were more likely to be women, as planned by the shelter, and were more likely to be black or Hispanic.

In general, then, the inner city location of Woods-Mullen had both advantages and disadvantages: employment, services and social supports were more readily available, but so were alcohol and drugs.

RECOMMENDATIONS

1. Provide support in the shelter for those who have experienced abuse as children or while homeless. Abuse is very common and is associated with a wide range of other problems.

2. Literacy help is needed by about one in five guests.

3. Providing AIDS education should be a high prioirty. AIDS has become a very great concern among shelter guests, with the importance of AIDS education ranked second only to help with finding a place to live. Further education also should lessen resistance to prevention activities: Although many were aware of prevention activities at the shelter, about one-third were not comfortable with these activities.

4. Substance abuse counseling continues to be a priority need, and should be emphasized particularly among guests using the Woods-Mullen shelter.

5. Drug abuse should be given as much attention in shelter program planning as alcohol abuse. These interview results identified a much higher level of drug abuse than had been found in case manager intake interviews.

6. The location of Woods-Mullen Shelter seems conducive to employment outside the shelter and to maintaining service and social connections. Some of these benefits might also be available to Long Island guests if day programs were increased at Woods-Mullen.

7. The shelter can be useful points for restoring some stability to homeless persons and for delivering needed services. Satisfaction with both shelters was high.

8. Efforts to help guests move into regular housing should be given the highest priority. The desire for "a place of my own" was seen as more important than any other service and was even more widespread than satisfaction with the shelters.

9. Legal counselors should be made available on a trial basis at the shelter. Law school interns might play a constructive role. Legal problems in the past and at present were very common, and reflected in part problems of substance abuse and mental illness.

10. Efforts to increase support from guests' natural support systems should be expanded. Family and friends were sources of support to many; with assistance, they might be able to help resolve the homelessness problem.

11. Homelessness prevention programs should be funded and publicized in the high risk neighborhoods: Dorchester, Roxbury, the South End.

12. The design of the Woods-Mullen Shelter should be altered to increase privacy between beds.