This is not my field of research either, but a Norwegian government
foundation for architecture, design and urban planning (Norsk Form) 
recently commissioned Norwegian Knowledge Center for the Health Services
(NOKC) to map and describe research in the field of health effects of the
physical environment. 

NOKC summarizes and disseminates evidence concerning the effect of
treatments, methods, and interventions in health services. When searching
for research that measures effect, they will look for study designs most
suitable for this purpose: randomized controlled trials. They focused their
search in the major health research data bases. 

Since the report (attached) is in Norwegian, I'll translate a bit of it here
including the search strategy. In the report, there is a one page english
summary. The tables of studies found and the reference list can also be
helpful even without knowledge of norwegian language.

From the english summary:
We systematically searched for controlled effect studies and systematic
reviews in international scientific databases, selected studies and
systematic reviews according to preset criteria, categorized the included
studies into sub-themes and summarised the reported results narratively in
tables. We did not assess the quality of the studies.

We categorised the included studies (n=65) and systematic reviews (n=10)
into the following sub-themes: interior decoration/furnishing (n=12), room
design (n=4), nature; plants, animals, sunlight (n=10), preventing falls and
wandering (n=5), lighting (n=6), noise reduction (n=7), multisensory
stimulation (n=5), distraction (n=21), miscellaneous (n=5). We found most
studies in the distraction category, but there are also studies that
measure the effects of sunlight in patient rooms, the effects of
refurbishing departments in both somatic and psychiatric hospitals, and the
effects of noise reduction strategies by adapting the physical environment.
The intervensions that are described in the interior decoration/furnishing
category are complex, and often include several different measures at the
same time.

The literature search did not retrieve evaluations of the effect of
artificial lighting for adults, or of the effect of wayfinding interventions.

In conclusion, there are controlled evaluations of the health effects of the
physical environment. Such effects may be measured through controlled
trials, and may contribute to a more relevant and reliable image of how
architecture and design may be used as a curative means. Experimental
research designs should to a greater extent be applied in this field of
research.

From the method section:
Search question: Is there research that investigates the effects of
architecture and design on life queality and health for patients at
hospitals, nursing homes or other institutions?

Population: patients at hospitals, nursing homes or other institutions
Intervention: architure or design-related actions (lighting, color, interior
solutions, etc.)
Outcomes: effect on health (number of bed-days, use of pain killer, grade of
stress, improved sleep quality, etc.) or life quality
Study design: randomized and/or controlled trials, systematic reviews

The followeing databases were search in november/december 2006:
Cochrane Library
OVID MEDLINE
OVID EMBASE
OVID PsycINFO
CSA Sociological Abstracts
OVID Webspirs IBSS
SveMed+
InformeDesign
SINTEF Byggforsk

Material found on the following sites was also reviewed:
The Design Council (UK)
http://www.design-council.org.uk
The Centre for Health Design Research (US)
http://www.healthdesign.org/

___

For more details about studies found, see attached pdf.
or download from the site: 
http://www.kunnskapssenteret.no/index.php?artikkelid=948&back=1


Sarah Rosenbaum
designer/senior advisor
Norwegian Knowledge Center for the Health Services