Skip to content

What do the academic studies say about community-based care?

August 27, 2012

A review of the literature

By David Kassel

Executive Summary

In the past two decades, there has been a major shift around the country in care for the intellectually disabled, as tens of thousands of people have been moved from larger congregate facilities to community-based group homes.

(This process has generally been referred to as “deinstitutionalization,” and I’ll use that term here despite the negative connotation of the  word “institution.”)

This shift in care, accompanied by a large-scale privatization of residential and other services, has engendered much debate among policymakers and advocates for the disabled.

That debate has played out for decades in Connecticut, picking up steam with the closure in 1993 of the Mansfield Training School by then Governor Lowell Weicker, and the filing of a lawsuit in federal court in 1996 to close the Southbury Training School and move its residents into a mix of public and privatized facilities.  Today, STS remains open, but many of the families and guardians of the residents there, who contend STS provides excellent care, are concerned that the state is moving to close it as quickly as possible.

In a recent letter to the STS Home & School Association, Terrence Macy, commissioner of the Department of Developmental Services, sought to reassure the families and guardians about leaving STS, stating, among other things, that:

[A] … significant body of rigorous studies has consistently concluded that persons who have transitioned out of institutions in fact have benefited from placement in the community.

Has the deinstitutionalization effort been an unqualified success, and does academic research support that claim?   

I reviewed academic research on the topic of deinstitutionalization of the intellectually disabled.  Much has been written in scholarly journals on this topic in the past 20 years, and I think it would be safe to say that the writers of those articles have given out both good and bad grades on the impact of deinstitutionalization on the well-being, health and longevity of the people involved. 

Specifically, the studies I reviewed found that:

  • Many people who were deinstitutionalized did indeed experience improvements in the quality of their lives and adaptive skills after being transferred to community-based group homes.

However, many of those same studies reported that:

  • Many deinstitutionalized people have encountered a lack of adequate medical care in the community system, and quality of life and adaptive skills often did not improve for those who had the most profound levels of disability or who were older.

I’ll first discuss the studies I reviewed on the well-being of deinstitutionalized persons and then will discuss an academic controversy over a study that found a higher death rate among people who were deinstitutionalized in California than what would have been expected in the institutions they were moved from.  (A list of the articles I reviewed can be found at the end of this post.) 

Quality of life studies on deinstitutionalized persons

One of the more comprehensive reviews I was able to find of well-being or quality-of-life studies of deinstitutionalized persons was a 2011 article by Monali Chowdhury and Betsey Benson in The Journal of Policy and Practice in Intellectual Disabilities.  

Chowdhury and Benson reported that their review indicated that relocation from institutions to community-based settings had a “general positive impact on life quality of participants.”   

At the same time, Chowdhury and Benson stated that many of the studies they reviewed found that “improvements were most prominent shortly after the move and plateaued after a year.”  While some studies showed gains in adaptive behavior in former institutional residents who had been moved to the community, they wrote, other studies “have shown neither significant increase in adaptive skills nor the maintenance of modest gains when moving to community group homes.” 

Chowdhury and Benson also reported that in a number of studies, “participants continued to have low levels of community integration,” and there was “some evidence that healthcare needs might not be adequately met in the community.”

A 2007 article in Current Opinion in Psychiatry by J. Beadle-Brown et al. contained the following summation of the academic literature on deinstitutionalization of the intellectually disabled:

Work on deinstitutionalization continues to show that outcomes are better in the community than in institutional care, but recent papers highlight that there is more to deinstitutionalization than just hospital closure.  Just moving people out of institutions into community settings does not bring about automatic improvement in quality of life in terms of choice and inclusion as well as self-identity and access to effective healthcare and treatment. This is especially true for people with more severe intellectual disabilities as well as complex needs such as challenging behaviour.

Among the other studies I reviewed that drew similar conclusions that transfers from institutional to community care were often but not always successful was a 2010 overview of research by Patricia Noonan Walsh et al. in The Journal of Policy and Practice in Intellectual Disabilities.

Citing what they termed “post-deinstitutionalization” studies that they had reviewed, Noonan Walsh and her colleagues stated that these studies provided “consistent evidence for greater choice, self-determination, and participation in community-based activities in smaller settings, but no evidence for greater physical health or material well-being…” 

In addition, a 1998 study in the journal Mental Handicap Research by Andrew Bratt and Rob Johnston of five intellectually disabled adults who had left a hospital to live in a “bungalow” in Exeter, England, concluded that while the staff considered the group as whole to be better off after the move:

…little evidence of integration being achieved within the local community was recorded; and there was little or no evidence of individuals being supported in ways that increased their competence, particularly in relation to participating at a simple level in routine domestic tasks.

Perhaps one of the reasons for the discordant findings among many researchers over the impact of deinstitutionalization is the difficulty in measuring that impact.  That difficulty may have contributed to the disagreement among many policymakers and advocates as to whether deinstitutionalization has been beneficial or not.

Chowdhury and Benson noted that many of the studies they reviewed on deinstitutionalization relied on interviews with intellectually disabled persons to determine quality-of-life differences between institutions and group homes.  However, other studies have found, they stated, that these interviewing techniques can be compromised by “response biases,” which “can seriously damage the reliability and validity of self-report data on subjective QoL (Quality of Life) in individuals with ID (intellectual disabilities).” 

One 2000 study, for instance, by Jonathan Perry and David Felce in Mental Retardation, reported that two thirds of the respondents in a sample of 154 adults with intellectual disabilities were either unable to respond to questions in a structured interview about their quality of life or exhibited response bias.   The authors noted that the most common form of response bias among people with intellectual disabilities is “the tendency to acquiesce, that is, to say yes to questions regardless of their content” (emphasis in the original).” 

Chowdhury and Benson further noted that given those problems with response bias, “a knowledgeable proxy,” such as a staff member or parent, is often asked to respond on behalf of an intellectually disabled person to quality-of-life survey questions.  However, they stated using such proxies is “usually not seen as an accurate substitute” for measuring QoL of people with ID.”

The California mortality study controversy

A major area of disagreement among academic researchers appears to center around attempts to determine whether moving intellectually disabled persons out of institutions causes them to die earlier than they would have had they stayed in those facilities.  This disagreement as well has spilled out into the policymaking and advocacy worlds.

I was recently criticized, for instance, by a leading opponent in Massachusetts of developmental centers for the intellectually disabled for citing a one such mortality study on deinstitutionalization  that had been done on residents who were moved from institutions to community group homes in California in the 1990s.

In a recent blog post  calling for saving the Glavin Regional Center in Massachusetts from closure, I wrote that a  2005 paper by Robert Shavelle, David Strauss, and Steven Day in The Journal of Data Science, reported that the authors had analyzed data  on more than 1,800 children and adults who had been deinstitutionalized in California between 1993 and 1996.  The authors found that as of 1999, there were 81 deaths among those who had been transferred from institutions to community-based group homes.  Using statistical regression, Shavelle et al. reported that the death rate represented a 47 percent increase in mortality over that expected in institutions.

Shavelle and Strauss concluded that the increase in deaths appeared to reflect less intensive medical care and supervision available in the community.

I wrote the post on behalf of the Massachusetts Coalition of Families and Advocates, Inc. (COFAR), a nonprofit organization that I consult for and that advocates on behalf of the intellectually disabled and their families.

COFAR, along with Strauss and Shavelle, were criticized by Gary Blumenthal, president of the Massachusetts Association of Developmental Disabilities Providers (ADDP), who wrote in a comment (under the user name “dcjayhawk”) to my post that:

COFAR…attempts to frighten families by telling them of higher mortality rates in community programs versus ICF (Intermediate Care Facility) programs by citing a flawed 1993-1998 California mortality study whose methodology and findings have been discredited by multiple expert researchers…

Blumenthal went on to cite four studies that he claimed discredited the work of Shavelle and Strauss.  (More about that below.)  The ADDP, which represents nonprofit operators of group homes in Massachusetts that contract with the state Department of Developmental Services,  supports the closure of all remaining developmental centers in Massachusetts and the transfer of all of their residents to community-based group home programs.

The California mortality research of Shavelle  and Strauss was also criticized by Macy, the Connecticut DDS commissioner.  In the same letter to the Southbury Training School Home & School Association referred to earlier, Macy maintained that:

The California study is well over twenty years old.  Its research methodology of contrasting deaths that occurred in the community to “those expected in institutions” is no longer a viable research methodology.  I would doubt this research design would even be accepted for publication in today’s research periodicals.  That study has been replaced with a large number of studies that have been conducted throughout the country over a long period of time.  Many of these studies are longitudinal so by design they follow persons who have moved over extended periods. 

Have the Shavelle study and earlier versions of it by Strauss really been discredited and are they really over 20 years old?  

In answer to that second question, I would note that Strauss and co-author, Theodore Kastner, first published findings of a higher mortality rate among deinstitutionalized persons in California in 1996 in The American Journal on Mental Retardation.   The article by Shavelle and Strauss, which I cited in my BMG blog post, was an update of the original study, and was published in 2005 in The Journal of Data Science.  Thus, even Strauss’s earliest findings do not appear to be more than 16 years old, and the work done by him and his colleagues on this topic was accepted by a peer-reviewed periodical only seven years ago.

Strauss’s research website lists a total of 10 studies and one commentary on mortality issues associated with deinstitutionalization that he co-authored with a total of eight colleagues.  Strauss himself is professor emeritus of statistics at the University of California, while Shavelle has been a visiting assistant professor at the University of California and has a Ph.D. in statistics.  Kastner, a physician, has held associate professorships at the New Jersey Medical School and the Columbia University College of Physicians and Surgeons.

I reviewed all four of the studies cited by Blumenthal of the ADDP as having discredited Straus’s methodology and findings.  What I found is that, as with the studies on the quality of life of intellectually disabled persons, studies on mortality rates of people who have been deinstitutionalized often reach differing conclusions.

I also found that while three of the four studies cited by Blumenthal did dispute the mortality findings of Strauss and his colleagues in the 1990s in California, two of those three were largely respectful of Strauss’s work.  (More about the exception below.)  Strauss and his colleagues, as far as I can tell, have not been discredited.  The fourth study cited by Blumenthal as having discredited Strauss’s California mortality studies actually did not directly address that topic at all. 

I would also note that each of the studies cited by Blumenthal that did dispute the mortality findings of Strauss and his colleagues was done by researchers who disclosed consulting relationships with state agencies that had or were seeking to close developmental centers.  In my view, those relationships raise questions about the independence of those studies.

As noted, only one of the articles cited by Blumenthal was actually overtly critical of Strauss et al. and their methodology.  That critical article, titled “Commentary on the mortality issue,” by James Conroy and Miriam Adler, was published in the journal Mental Retardation, in October 1998.

In the commentary, Conroy and Adler cited a 1995 study they undertook on the transfers of residents from the Pennhurst Developmental Center in Pennsylvania, and stated that they had found a “sharply decreased risk of death when people moved from institution to community.”  They termed Strauss’s 1996 finding of a higher death rate in the community system “utterly false” and compared it to the discredited discovery of cold fusion.

It should be noted that Kastner returned the favor, critiquing Conroy’s methodology in a number of papers, including a 2000 article  in Mental Retardation concerning Conroy’s Pennhurst closure study.  Kastner also co-authored a 2006 article with Kevin Walsh in the same journal, titled “The Hissom Closure in Oklahoma: Errors and Interpretation Problems in Conroy et al. (2003).”

In that 2006 article, Walsh and Kastner charged that the research of Conroy and his colleagues had been “seriously compromised by data collection problems and discrepancies between reported findings and those obtained when the analyses were replicated.”

Several of Conroy’s reports were also criticized in the 1990s by former staffers of the California Department of Developmental Services, according to the May 1996 issue of Outreach, a publication of the California Association of Psychiatric Technicians.  The journal article, titled “Conroy Reports Controversy,” also stated that Philip Bondy, who is a former chair of Yale’s Department of Internal Medicine and an STS Home & School Association Board member, was critical of a study done by Conroy on the outcomes of the closure of the Mansfield Training School in Connecticut.  

Bondy emailed me that his criticism was based on Conroy’s reliance in the Mansfield study on quality-of-life interviews with attendants in community-based group homes as proxies for former Mansfield residents.  “There is no reason to believe that the attendants were unprejudiced,” Bondy wrote, “since they were employed in providing community care and might reasonably expect to give favorable answers.  Although Conroy quoted statistics in  support of his conclusions, I believe that they were based on faulty data,” he added.

Bondy is a former editor in chief of both The Journal of Clinical Investigation and The Yale Journal of Biology and Medicine.

Conroy’s resume  indicates that he authored or co-authored hundreds of reports on the well-being of deinstitutionalized residents, most of which appear to have been submitted to state agencies running ICFs and community-based facilities or to entities such as the Institutional Closure Commission of Ohio.  I think that may raise some questions about his independence in researching this issue.

A second article cited by Blumenthal as having discredited the work of Strauss and his colleagues on mortality rates in California was a 1998 study by Kevin O’Brien and E.S. Zaharia.  The study, titled “Recent mortality patterns in California,” was published in the same issue of Mental Retardation in which the Conroy commentary was published.

The O’Brien and Zaharia study found no increased risk of death in the community for the same population that Strauss and his colleagues had studied.   But far from discrediting Strauss’s results, O’Brien and Zaharia stated that the reasons for the discrepancies between the two studies were “unclear.”  They acknowledged that due to validity problems with databases used to analyze the number of deaths, “there may have been quite a bit of contamination in all of our analyses.”

In a separate commentary in the same issue of the journal (titled “Is it Life Threatening to Live in the Community?”), O’Brien and Zaharia stated:   “We believe that the suggestions made by Strauss et al. (1998) (about closely monitoring initial placements of residents in community settings) should receive thoughtful debate.”

O’Brien and Zaharia disclosed in their papers that they were consultants to the California Department of Developmental Services.

A third study cited by Blumenthal as discrediting the Strauss studies was done by Paul Lerman, Dawn Hall Apgar, and Tameeka Jordan and published in Mental Retardation  in August 2003.  This study did not find evidence that moving residents out of a developmental center in New Jersey and into the community caused more deaths.  However, the authors concluded that the difference between their results and Strauss’s findings could be due to variations in policies in the two different states on deinstitutionalization or to differences in population characteristics. 

Lerman et al.  also noted several limitations to their own study, including limitations on the reliability of their behavioral variables.   The study was hardly an effort to discredit Strauss and his colleagues.

The Lerman study, by the way, was funded by the New Jersey Division of Developmental Disabilities.

The final study cited by Blumenthal as discrediting the Strauss studies was actually concerned with the well-being of people transferred from nursing homes to community residences.   The 1998 study, by Tamar Heller, Alan Factor, and Kelly Hsieh, in The American Journal on Mental Retardation, mentioned Strauss’s research in passing and was not in any way critical of it.

In a defense of their work, published in the same 1998 issue of Mental Retardation as the Conroy and O’Brien and Zaharia articles, Strauss and Shavelle maintained that they were not attempting to evaluate “the merits of institutional versus community living,” and were not arguing that persons living in the community should move or return to a state institution.”  Instead, they said, “[o]ur concern was with the quality of health care and supervision currently available in the community.”

In that commentary, Strauss and Shavelle cited a Congressional subcommittee report, newspaper reports on avoidable deaths in California group homes, and a critical report on the community system in California by the federal Health Care Finance Agency as buttressing their findings. 


This review of the academic literature on institutional versus community-based care for the intellectually disabled is admittedly far from exhaustive.  But I think most researchers in this field would probably agree that I’m correct in concluding at least one thing about the research that has been done on this topic:  That is that there is no clear consensus among scholars that either system of care is better in all respects for all intellectually disabled people. 

Both of the advocacy organizations that I work for believe that the community-based system is appropriate and beneficial for the majority of people with intellectual and developmental disabilities.  At the same time, both organizations also maintain that for a portion of this population with severe and profound levels of disability as well as those who who are older or have complicated medical conditions, institutions such as STS continue to play a vital role in their care.

Both COFAR and the Southbury Training School Home & School Association are nonprofit organizations that advocate for families and guardians of intellectually disabled residents of state-run facilities.  Many of those facilities have been targeted for closure by state administrators who, in many cases, appear to believe that all intellectually disabled persons should live in the community system.

I hope this survey of the literature on this topic will persuade at least some of those administrators as well as legislators, policymakers, and others that those in the academic community, who have looked at this issue objectively and scientifically, have not reached a consensus that the community is better for everyone.  Instead, the scholarly literature on this topic appears to support the contention that “one size doesn’t fit all.”

References for this post:

Beadle-Brown, J.; J. Mansell; and A. Kozma, “Deinstitutionalization in intellectual disabilities. Current Opinion in Psychiatry, 2007, 20(5):437-42.

Bratt, A. and R. Johnston, “Changes in life style for young adults with profound handicaps following discharge from hospital care into a ‘second generation’ housing project,” Mental Handicap Research,  1998, 1(1): 49–74.

Chowdhury, M., and B. Benson, “Deinstitutionalization and Quality of Life of Individuals With Intellectual Disability: A Review of the International Literature,” Journal of Policy and Practice in Intellectual Disabilities, 2011, 8(4): 256–265.

Conroy, J., and M. Adler, “Commentary on the mortality issue,” Mental Retardation, 1998, 36(5).

Cullen, C.;  M. Whoriskey, K. Mackenzie, W. Mitchell, K. Ralston, S. Shreeve, and A. Stanley, “The effects of deinstitutionalization on adults with learning disabilities,” Journal of Intellectual Disability Research, 1995, 39(6): 484–494.

Heller, T.; A. Factor, and K. Hsieh, “Impact of age and transitions out of nursing homes for adults with developmental disabilities,” The American Journal on Mental Retardation, 1998, 103(3).

Kastner, T., “On the Need for Policy Requiring Data-Sharing Among Researchers Publishing in AAMR Journals: Critique of Conroy and Adler (1998),” Mental Retardation, 2000, 38(6): 519-529.

Lerman, P.; D. Apgar, and T. Jordan, “Findings of a Controlled Research Design in New Jersey,”  Mental Retardation, 2003, 41(4): 225-236.

Noonan Walsh, P.;  E. Emerson, C. Lobb, C. Hatton, V. Bradley, R. Schalock, and C.  Moseley, “Supported Accommodation for People With Intellectual Disabilities and Quality of Life: An Overview,” Journal of Policy and Practice in Intellectual Disabilities, 2010, 7(2): 137–142.

O’Brien, K., and E.S. Zaharia, “Recent mortality patterns in California,” Mental Retardation, 1998, 36(5).

O’Brien, K., and E.S. Zaharia, “Is it life threatening to live in the community? Commentary,” Mental Retardation, 1988, 36(5).

Perry, J. and D. Felce, “Subjective and Objective Quality of Life Assessment: Responsiveness, Response Bias, and Resident: Proxy Concordance,” Mental Retardation,  2002, 40 (6): 445–456 

Shavelle, R.; D. Strauss, and S. Day, “Deinstitutionalization in California: Mortality of Persons with Developmental Disabilities After Transfer Into Community Care, 1997-1999,”  Journal of Data Science, 2005, 3(4): 371-380.

Strauss, D. and T. Kastner,  (1996). “Comparative mortality of people with mental retardation in institutions and the community,” American Journal on Mental Retardation, 1996, 101(1): 26-40.

Strauss, D. and R. Shavelle, “Policy Implications of Mortality Research: What Can We Learn From the California Mortality Studies?” Mental Retardation, 1998, 36(5).

Walsh, K., and T. Kastner, “The Hissom Closure in Oklahoma: Errors and Interpretation Problems in Conroy et al.,”  Mental Retardation, 2003, 44( 5).

Weismann, C., and K. Hearn, “Conroy Reports Controversy,” Outreach, May 1996.


From → Uncategorized

One Comment
  1. Anonymous permalink

    Another astute, well-balanced and exhaustively researched analysis by David Kassel. Thank you once again for helping those of us who must make this incredibly difficult decision see things more clearly. Who can argue against STS remaining open in one form or another for the most unfortunate? Jim Carrington

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: