Designing an evidence-based research project: The Children's Hospital collaborative approach
H+L Architecture, in association with Zimmer Gunsel Frasca Architects, started designing a new facility for The Children's Hospital (TCH) at the former Fitzsimons Army Medical Center in Aurora, Colorado, in 2003. This not only presented an opportunity to create a jewel in the world of healthcare, but it also inspired an important research study on how a healthcare environment can positively affect its patients, families, and staff.
During TCH's design process, H+L Architecture became interested in conducting a research study comparing aspects of the old and new hospital environments. This research study used evidence-based design; a relatively new concept in the construction industry. An evidence-based approach to design is based on the same approach that has been used for clinical care. Evidence-based medicine is a methodology where clinical care is based on the preponderance and strength of the evidence on which the practice is founded. The evidence is data- and outcome-driven. When evidence-based design is used to design a healthcare facility, the approach is based on research that links elements of the built environment with patient and staff outcomes. As Hamilton defined evidence-based design, it is “a process for the conscientious, explicit, and judicious use of current decisions, together with an informed client, about the design of each individual and unique project” (Hamilton, 2003).
Designers' use of evidence-based design in healthcare architecture has traditionally been limited due to a lack of existing research studies to use as a guide. Previously, designers and hospital staff relied on methods including healthcare benchmark tours and seeking the advice of experts in the field, as well as review of the literature. In fact, during the TCH schematic and design phases, one question was repeatedly asked by the users and the executive administration: What does the literature say? As the design team reviewed the existing literature they discovered there was a dearth of information on healthcare design's affect on patients and staff.
H+L had previously participated in research with the University of Colorado Hospital and was eager to supplement that initial study with additional data. H+L approached TCH in the fall of 2004 with a proposal to conduct research on how design decisions achieved the operational goals expressed during design development. Additionally, TCH would be able to compare data between the previous and new facility, adding a second dimension to the research. Ideally, research hypotheses should be developed by executive leadership during the conceptual phase, long before actual design starts. When H+L approached TCH, the design portion of the project had been completed and most decisions had been made. Nonetheless, given the project's unique nature as one of the largest healthcare project in the nation at the time, as well as the ability to compare pre and post move, there were still valuable insights to be gained. Subsequently H+L, ZGF, and TCH formed a research collaborative as a Pebble Partner in the fall of 2007.
Research proposal development
As an initial step in creating the research proposal, the research team at TCH had discussions with hospital staff regarding the critical design elements that elicited the most discussion during the design development meetings. In addition, the architects and TCH staff reviewed the “Guiding Principles,” developed by TCH's executive leadership, which included concepts such as Timeless/Enduring, Clarity/Orientation, Family-centered, and Not Trendy. Two final concepts that were instrumental in proposal development were investigating if TCH had received a good return on its investment by utilizing a cost-benefit analysis of the designed building, and the tenet that the building should reflect the operational goals of the organization. Using these four approaches to the proposal, it was decided to study the design through a specific concept, such as decentralized charting, new adjacencies, communication pathways (silent hospital), and the effect that concept has on daily operations.
Once the research team and executive leadership weighed in on these tenets, the next step in developing the study was to develop specific aims. The aims chosen by TCH addressed staff and family satisfaction and patient outcomes. The major areas that would be evaluated within those outcomes were decentralized charting alcoves, light and noise, patient and staff flow/efficiency, family and staff satisfaction with parking, wayfinding, and safety.
The next step was to develop the research methodology. During this process H+L and TCH considered who would be surveyed, what data to collect that measures the specific aims, as well as how the collection would be implemented. Data that lent itself specifically to evaluating patient, family, and staff outcomes included differences in nosocomial infection rates, staff ratios, staff turnover, staff and family satisfaction with layout, family respite space and amenities, security and wayfinding, etc. The team decided that the data would be collected using a pre-/post-move evaluation approach. Data would be collected at the previous hospital and the new.
Development of data collection instruments was the next task. Measuring staff and family satisfaction through a survey was investigated first. This type of data collection instrument needed to be focused specifically on the built environment. After reviewing several different instruments that measured building design and staff/family satisfaction the team developed a survey instrument specific to the project. Developing this survey took over a year and required a significant amount of time from both TCH staff and H+L designers.
Staff and family satisfaction
The survey was piloted in 2006 with both staff and family members. After establishing validity and reliability of the staff, family, and Environmental Services (ES) surveys, we were able to use these instruments to evaluate staff and family satisfaction pre-move and post-move. The survey was translated into Spanish for both the ES staff and for family members that preferred to complete the survey in Spanish.
In addition to these surveys, the hospital Press Ganey industry benchmark for patient/family satisfaction was used to obtain outcome data on the perception of satisfaction with the built environment. Using this historical information on family satisfaction augmented information obtained from the built environment survey. One complication that arose with the hospital family satisfaction data was a decision by the hospital to change vendors after the first year in the new hospital. TCH used Press Ganey, Inc., family satisfaction surveys for pre- and postoccupancy for year one. A change was made to Professional Research Consultants, Inc., surveys for year two.
To fully vet TCH's goals for measuring staff and infection control specific aims, meetings were set up with TCH's Nursing Administration, Human Resources, and Infection Control Departments to develop measurement instruments specific to their department.
The team was able to use ongoing data sources for nursing and staff turnover and nursing ratios (proxy for staff efficiency in the new building). The current caregiver/patient ratio was used for comparison between the previous and new hospital. The Human Resources Department shared its staff and nursing turnover and vacancy statistics.
The Infection Control Department data was difficult to correlate to the built environment as there were so many variables affecting the data that did not relate to the building design, such as community infection types and rates. In lieu of nosocomial infection rates, hand-washing data was reviewed. However the data gathering methods were different before and after the move to the extent that correlation was difficult to establish. The rate for 2006 was 76.25%. In 2007, three months after the move to the new hospital, the rate was 80.2% before patient contact and 92% after. In December 2008, fifteen months after occupancy, the rate before patient contact was 96% and after was 97%. Even with the correlation issues, the increase in these rates in 2008, post move, were impressive when compared against the CDC mean baseline of 5%-81% (overall average 40%) (CDC Guidelines for Hand Hygiene, 2002) or Ulrich who found compliance rates in the area of 15%-30% as typical (Ulrich 2004).
Light and noise
The light and noise measurements used the “point-in-time” approach to measure the levels of light (foot candles) and noise (decibels). The measurements were taken at the same time of day, four times a year, to ensure consistency for the light readings and capture difference related to each of the four seasons. Pre- and post-move measurements were made but establishing comparable areas in the new and previous hospital was difficult since the operations and building designs between the two hospitals were dramatically different. An acoustical engineer from Johns Manville, a Pebble Project corporate member, assisted in the noise readings. The light measurements were done by TCH staff. The results of the light and noise measurements proved interesting when correlated with the staff surveys and the anecdotal comments that were recorded from staff by administration. While the actual noise readings were well within industry standards and lower than the previous building, the staff had a negative impression of the noise levels in the patient care areas. This study of light and noise data within the built environment is an area of focus for future research using a geographic mapping tool.
Though comparing patient/staff flow pre and post move was an area of interest, the research team was unsuccessful in developing a good measurement approach. This is a topic for future research as many design decisions are based on assumptions of how staff will operationalize building design. Testing to see if the assumptions that an individual facility has about clinical flow and staff function held true is needed when operational flow and clinical approaches to care are impacted by the building design.
After many months of proposal and instrument development, the next step was to obtain Institutional Review Board approval (IRB). Developing processes for distributing and collecting the surveys as well as estimating the number of subjects to be surveyed were among some of the details to be worked out before the project could be submitted. After IRB approval was obtained, a communication plan was developed. In many cases the communication plan was combined with the implementation plan. For example, as the survey began to be implemented, researchers attended numerous staff meetings to explain the research focus and elicit staff participation. The specific aims including light and noise, use of design features etc. were explained in question and answer sessions with departments. In order to maximize completed surveys and explain the research goals, formal presentations were conducted. Several key departments including nursing units, senior management, and the hospital planning department were targeted for these presentations. In addition, articles were published in TCH's internal newspaper, TCH News, as well as other TCH publications such as the donor periodical, Impact.
The majority of the built environment study was completed in 2008. Surveys were distributed pre and post move. The other data, staffing, hand washing, and noise and light readings were collected pre and post occupancy. The only additional data collection may be one short and targeted staff survey. Even employing a careful and targeted research approach with selective and specific aims using customized data instruments the volume of data collected was massive. The research team is in the process of final analysis and will be publishing the results soon.
The next step for the team after analysis and publication is to ensure that the results are integrated into TCH's planning and future building. In addition the architect members of the research team will be able to use these results when designing their future medical facility projects. The researchers put forth these results to add to the body of knowledge available to all in the healthcare community. There are several follow-ups that are in process, including a more in-depth study of the effect of a new and larger building on full-time equivalents (FTEs), as well as a short and targeted staff survey.
The project started as a research collaborative between The Children's Hospital (TCH) staff and H+L with the in kind labor effort being donated by the staff and H+L employees and underwritten by H+L architecture. Originally the effort included Susan Koch Zacharakis and Anne Marie Kotzer from TCH and Mary Raynolds, Michelle Rademacher, and Rob Davidson. This small group worked to develop a list of potential research topics to look at pre- and post-move conditions. The original group also developed the staff, Environmental Services Employees, and family survey tool; got approval from the IRB to survey the families; and translated the Environmental Services and family surveys into Spanish in order to expand the number of survey responses. The pre-move survey and data were gathered primarily by TCH staff. After completion of construction, activation, move-in, and a subsequent change in hospital administration, TCH expressed an interest in making the ongoing research effort a Pebble Project and TCH joined the Pebble Project with financial support from H+L and ZGF. The continuation of the research effort including the post-move physical data gathering and family and staff surveys were done under the umbrella of the Pebble Project with the survey data gathering done by in kind labor by TCH staff and the physical data gathered with assistance by Johns Manville as a corporate Pebble member.
We had many challenges through our proposal development and project implementation. Several were associated with the survey process. One was the challenge of distribution of the family surveys, as well as the pickup of completed surveys. This was a challenge due to the random visiting patterns of families, language barriers, families living far from Denver, and the lack of a central repository for returning the surveys.
A second challenge with the survey portion was the ES survey. We chose to eliminate surveying the ES staff. This was a difficult decision as their experience and opinion in regards to the finishes were highly valued. However it was found that staff turnover and reluctance to voice their opinion decreased the usefulness of their input. ES employees provide a unique viewpoint and future researchers should obtain their opinion even though it may be difficult to elicit.
Another limitation to our study occurred as a result of the hospital's decision to change satisfaction survey vendors during the research project. This change resulted in different questions and measurement methods but to eliminate the post-move “honeymoon” period it was important to include the second post-move year data even with the vendor change.
Correlating the noise and light measurements was a challenge as well when the new and old geography did not match. We chose spaces by the same name (e.g., PICU medication room), but the building configurations were often very diverse creating challenges in comparison.
Other limitations previously outlined included difficulty in developing a methodology to measure staff flow and travel distance and being unable to tie differences in nosocomial infection rates back to private versus semiprivate patient rooms.
The development, implementation, and analysis of research on the built environment are labor-intensive activities. Dedicated time is required from clinical and architectural staff, as well as consultation from research experts and other consultants, depending on one's research questions. The best scenario involves having a dedicated research team devoting their time exclusively to the project. The TCH/H+L team, which consisted of clinical and research TCH staff and designers from H+L, did not have the luxury of a full-time team but were able to complete their study while having the team perform research duties in addition to their daily job responsibilities. HD
Susan Koch Zacharakis, MSHA, RN is former Director of Clinical Planning, and Anne Marie Kotzer, PhD, RN, CPN, FAAN, is Senior Nurse Researcher at The Children's Hospital, Aurora, Colorado. Fred Buenning, AIA, Principal, and Mary Raynolds, Assoc. AIA are both with H+L Architecture.
- Guidelines for hand hygiene in health-care settings: Recommendations of the Healthcare Infection Control Practice Advisory Committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force (2002).Retrieved July 7, 2009. http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5116a1.htm
- Hamilton K. (2003, November). The four levels of evidence-based design practice. Healthcare Design, 18-26.
- Pebble Project, Retrieved on May 1, 2009, http://www.healthdesign.org/research/pebble/
- Ulrich R., Quan X., Zimring C., Joseph A., & Choudhary R. ,(2004) The role of the physical environment in the hospital of the 21st century: A once-in-a-lifetime opportunity. Center for Health Design, Concord, CA.
Healthcare Design 2010 May;10(5):56-67