Skip to main content

Observation Studies: How Care Providers Spend Their Time in Long-Term Care Facilities

Case
By: & Published: 2017 | Product: SAGE Research Methods Cases Part 2
+- LessMore information
Search form
No results
Not Found
Download Case PDF

Abstract

This case study describes how we conducted an observational study to explore the activities of care providers in long-term care facilities and to determine whether variations existed within and across different roles, shifts, and facilities. We focus on how observational data of care activities can be collected in a manner that does not violate the privacy of residents living in the facilities. We discuss how the requirement for informed consent was managed given the ever-changing population of residents, staff, visitors, and volunteers. We also discuss how observation data can be collected, stored, and analyzed.

Learning Outcomes

By the end of this case, students should be able to

  • Have a better understanding of the benefits and challenges of observational studies
  • Have an awareness of how the experiences and background of researchers conducting observations can affect the data collected
  • Reflect on how an awareness of being observed may affect participants’ actions
  • Consider how to share study findings in a manner that is meaningful and simultaneously respects the anonymity of participants
Project Overview and Context

The aim of the research was to understand how care providers spend their time in long-term care facilities and to identify the types of activities in which they engage. In the Canadian province where the study took place, long-term care facilities are government regulated and operate according to a mandated staffing ratio that outlines required staffing levels. This ratio specifies both the number and type of staffing mix that are required to care for a given number of residents. For example, when this study was conducted, the standard staffing ratio was 3.1 hr of care per resident each day, composed of 40% unregulated Resident Aids (RA), 40% Licensed Practical Nurses (LPN), and 20% Registered Nurses (RN). At the time, there was concern by those working in the sector that such a staffing standard was inadequate given the complexity and often multiple care needs of residents. Although calls for additional staffing were made, there was no agreement on what an optimal staffing standard would be or any consensus on what specific changes in staffing were required (i.e., additional RAs, LPNs, or RNs). A more in-depth understanding on how care providers spend their time and contribute to resident care was recognized as a necessary component for identifying appropriate levels of staffing.

One unique aspect of this research was that data were collected using participant observation rather than the traditional structured or semi-structured methods. Observations of staff activities provided an objective and non-biased account of the phenomena of interest while allowing context to be considered, and in so doing, the researcher became the data collection instrument.

Our research team consisted of two RNs and an undergraduate university student who served as a research assistant. One of the RNs was a health care administrator, whereas the other was an academic. The RNs were responsible for developing the research design, analyzing the data, meeting with appropriate people at each study site, obtaining research and ethical approval, and orientating the research assistant to the data collection procedures.

Research Design

This research aimed to determine how care providers spend their time in long-term care facilities and to identify the types of activities in which they engage. A major component of the research was to understand how contextual factors such as the activities of other care providers and the particular shift being worked influenced staff activities. The observation of care activities provided a means to meet both of these goals. DeWalt and DeWalt (2002) contend that participant observation is ideal for studying processes and the organization of people, thus making it an ideal method to use to study staff activities.

Accessing Participants

To obtain a realistic representation of staff activities, we decided to collect data from a variety of long-term care facilities. Considering both financial resources and time, we purposively selected seven different facilities that offered some diversity in terms of facility design and age. Five of the facilities were considered new and had predominantly single resident bedrooms with small self-contained living and dining spaces. The other two facilities were older and had multi-resident bedrooms and large common public spaces and dining rooms. It was our belief that these facilities would allow us to make some comparisons across homes as well as provide an understanding of how contextual factors such as building design, number of residents residing within a facility, and the distance between residents might affect staff activities. In selecting the exact facilities to include in the study, we consulted our contacts within the long-term care sector. We intentionally wanted to avoid facilities with known labor relation issues or high staff turnover. We were also interested in approaching facilities with a reputation of being open to research and receptive to receiving feedback on their day-to-day operations.

After deciding which long-term care facilities to include, our team met with senior administrative personnel in each of the facilities to explain the purpose of the study and what participation would involve. It was during these meetings that we stressed the confidential nature of the study, and although we promised to share our findings with each participating facility, we also assured them that the identity of their facility and facility-specific data would not be shared with others. Further to this, we asked the administrator in each individual facility to select a unit considered to be a general nursing ward. In other words, it is a unit that did not offer specialized services to a subpopulation of residents such as a dementia, palliative, or respite care unit. We decided to exclude specialized units to minimize confounding variables and to maximize comparisons that could be made across facilities. This was considered an important decision because some facilities did not have any specialized resident units whereas others did.

Sample

Although our primary interest was to observe the direct care providers, we were aware that non-care providers would also be present in the setting during the periods of observation. Naturally, residents would be in the setting, but so would employees of each facility who were not care providers as well as family members, visitors, and volunteers. Recognizing that these individuals would help to contextualize the data being collected and perhaps allow for a follow-up study to be conducted, we decided to expand our sample to include these individuals.

Consent

Our research study was reviewed by six Research Ethics Boards (REBs), and permission to proceed was granted by all boards; three facilities shared the same REB plus the study was reviewed by our university’s REB. Obtaining consent from individuals in each of the areas of observation was a complex but necessary activity. This action involved informing people that their activities and actions would be observed, as well as of any risks and benefits of participation (DeWalt & DeWalt, 2002). The obtaining of consent for this study was particularly challenging due to the unpredictable number of family members and visitors who could potentially enter into the observation areas on any given day and time. To address this issue, we sought approval from the REB in each facility to employ an “assumed approach” to consent rather than a written consent for each participant. The “assumed approach” involved sending letters to all family members, volunteers, and staff in each facility to explain the nature, purpose, and dates of the proposed research. Meetings were held with care staff on the specific units where observations were planned to provide details of the study and answer any questions and/or address concerns. All individuals living, working, or potentially visiting a participating study unit during the period of observation were asked to notify a designated person of any objections to the study and/or being observed. In addition, during data collection, signage was posted in each care unit notifying and reminding people that the study was taking place. Nobody voiced any concerns about the study or about being observed during the data collection phase of the investigation.

The Observer

One of the most important aspects of conducting an observational study is being able to carry out effective observations. For our study, this included being mindful of the multitude of activities and people in a defined space at any given time. An important question we asked ourselves during the very early stages of the research was “what qualities and qualifications would be required of the individual(s) collecting data?” We were mindful of the fact that an individual with health care experience may have a greater degree of familiarity with the care environment than somebody without such experience. Having an individual with first-hand knowledge of the environment would also provide an opportunity to capture data on social nuances that took place. Conversely, we questioned whether, and in what ways, an observer’s familiarity and comfort in an area could lead to pre-determined assumptions about observed activities.

After considerable discussion and debate, we opted to utilize an observer who had no health care experience and/or familiarity in the environment under investigation. This decision required us to conduct an intensive orientation to the environment with the individual who was collecting the data (i.e., performing the observations). This included an overview of the various classifications of workers present in the long-term care settings and also an orientation for the observer with the physical spaces where observations would take place and some of the behaviors and/or activities that could be expected. Given the study took place in long-term care, it was also important to acclimatize the observer to the various noises and smells that they would likely encounter, which could be potentially distracting if unexpected. To accomplish this, we asked the staff working in each facility orientate the observer to the physical space where the data were being collected. Further to this, we also spent several hours in a long-term care unit not included in the study collecting data independently, but alongside the observer. This co-collection of data continued until the data collected by one of the researchers, who was a health care professional, corresponded with the data collected by the observer. This activity was considered necessary to help establish reliability of data collected by the trained observer.

Data Collection/Recording Observations

Observational data can eliminate biases that may arise during self-reports or retrospective descriptions of events as differences often exist between how people describe their actions and what they actually do (Lee, 2000). Different methods of data collection can be employed in observational studies, including direct observations and video/tape recordings. In some observational studies, participants are fully aware that they are being observed. In these cases, the observer may ask questions or seek clarification about what is being observed (Eika, Dale, Espnes, & Hvalvik, 2015; McCloskey, 2011). In other cases, the observer attempts to blend into the background in hopes of not disturbing or altering participants’ natural behaviors or activities (Mallidou, 2013). Conversely, participant observation can involve covert methods where participants are unaware that they are being observed. Although covert observations present many ethical issues, it has been argued that people are likely to alter their behavior if they are aware they are being observed (Spicker, 2011). The notion that people may alter their behavior based on being observed is referred to as the Hawthorne effect (McCambridge, Witton, & Elbourne, 2014). However, by minimizing interactions with people in the environment, the researcher can reduce the risk of the Hawthorne effect (DeWalt & DeWalt, 2002).

In our study, we decided to minimize the data collector’s interactions with those in the environment. To accomplish this, the observer collected data by walking an identical route repeatedly throughout each of the facilities during the periods of data collection. When asked, the observer responded to questions about the study, but otherwise did not interact with anyone in the setting. Each route was intentionally selected to allow observations of all locations and activities within each study site. The only exception was in areas that could potentially violate residents’ privacy, such as bathrooms and bedrooms, and so the data collector was prohibited from entering these areas. In cases where the door into these rooms was opened, the observer could look inside but could not enter.

Observations were entered into a mobile device each time a care provider was observed. These observations were entered as textual data that represented and included both the category of provider and the activity observed. The mobile device used in our study was programmed with Time Study RN software. This software program is relatively inexpensive and can be purchased and downloaded onto most computers. The software had the capacity to organize observations entered into pre-arranged categories or data sets. For our study, we programmed the mobile device so that data entered would automatically be organized into both categories and subcategories. Categories were broad groupings of data including the individual observed (care providers vs. resident), direct care activities, and indirect care activities. Each category was then further divided into distinct subcategories. For example, the category of direct care activities included activities such as resident care, medication administration, and feeding. The precise time of each observation was automatically attached to each data entry. Although unanticipated, having the exact time of each observation allowed us to make important inferences about the data. For instance, knowledge of when care providers were feeding residents allowed us to challenge prevailing assumptions about team work and scope of practice in each of the long-term care facilities. Although not part of our study, others have used the time of each data entry to determine average length of time for care providers’ encounters with residents (Mallidou, Cummings, Schalm, & Estabrooks, 2013).

Analysis

The information entered into the mobile device was transmitted directly into Excel 2010. Because of the variation in facility sizes, the number of staff working in each facility at any given time varied greatly. To avoid misrepresenting the time staff spent in each individual activity, we converted actual observations to proportions of observations. In other words, we looked at all of the observations we had for care providers and then calculated the percentage of these observations that were spent in each specific care activity. By analyzing proportions of time spent in activities, we were able to provide more meaningful comparisons across shifts and facilities.

Data were analyzed individually for each facility and cumulatively for all seven facilities. This allowed us to identify the uniqueness within each facility and to compare activities among all participating facilities. For example, we were able to compare differences in the time staff spent with residents across all facilities and then to further examine this with like-facilities (i.e., compare facilities with predominantly single bedrooms and differences between facilities with single vs. multi-resident bedrooms). Conversely, by combining all the data, we were able to compute an average from all observations, which each facility could then use to interpret their own results. For instance, the combined RA activities in one faculty showed 4% of time was spent in non-value-added activities. When taken in isolation, this figure may raise concerns; however, when compared with 23.7% observed in another facility and the 15.2% mean across all participating facilities, this figure elicits a very different reaction.

Conclusion

This case study highlights the many practical issues that researchers must address when conducting an observational study. First, the researcher must consider the challenges involved in obtaining consent to participate in a research study when it is unknown who will be in the setting. Second, the researcher must consider the best participant observation approach to take; should the observer interact with participants or merely observe their activities? Third, the researcher must determine which pieces of data the observer will focus on in the setting and how these data will be organized and later analyzed. Last, an important consideration that must be made up front is whether the data collector should be known or unknown to the setting and/or those being observed. Throughout this case study, we have explained how our team addressed the challenges inherent within observational studies and how these decisions helped us to explore the activities of care providers working in a health care setting.

Exercises and Discussion Questions
  • Discuss the pros and cons of conducting an observational study.
  • How could you obtain informed consent in an observational study where the individuals entering the setting are unpredictable?
  • Does the knowledge that one is being observed affect behavior? If so, are there techniques that can minimize this?
  • In an observation study, is it necessary for the observer to be familiar with the practice area where the study is taking place?
  • Would covert methods ever be justified? Do covert observational studies in public spaces violate any ethical principles?
References
DeWalt, K. M., & DeWalt, B. R. (2002) Participant observation: A guide for fieldworkers. Walnut Creek, CA: AltaMira Press.
Eika, M., Dale, B., Espnes, G. A., & Hvalvik, S. (2015). Nursing staff interactions during the older residents’ transition into long-term care facility in a nursing home in rural Norway: An ethnographic study. BMC Health Services Research, 15, Article 125. doi:http://dx.doi.org/10.1186/s12913-015-0818-z
Lee, R. M. (2000). Unobtrusive methods in social research. Philadelphia, PA: Open University Press.
Mallidou, A. A., Cummings, G. G., Schalm, C., & Estabrooks, C. A. (2013). Health care aides use of time in a residential long-term care unit: A time and motion study. International Journal of Nursing Studies, 50, 12291239. doi:http://dx.doi.org/10.1016/j.ijnurstu.2012.12.009
McCambridge, J., Witton, J., & Elbourne, D. R. (2014). Systematic review of the Hawthorne effect: New concepts are needed to study research participation effects. Journal of Clinical Epidemiology, 67, 267277. doi:http://dx.doi.org/10.1016/j.jclinepi.2013.08.015
McCloskey, R. M. (2011). A qualitative study on the transfer of residents between a nursing home and an emergency department. Journal of the American Geriatrics Society, 59, 717724. doi:http://dx.doi.org/10.1111/j.1532-5415.2011.03337.x
Spicker, P. (2011). Ethical covert research. Available from: http://openair.rgu.ac.uk

Copy and paste the following HTML into your website