A write-up of my research and design project done in collaboration with Memorial Sloan-Kettering Cancer Center.
A research and design project done in collaboration with Memorial Sloan-Kettering Cancer Center to investigate the period of rehabilitation that follows a bone marrow transplant and then design improvements for patients and medical staff.
The project features original research into the experience of a bone marrow transplant from the perspective of medical staff, administrative staff and transplant recipients. The findings from this research are grouped into six themes: The Patient’s Journey, Symptoms, The Whole Patient, The Importance of the Caregiver, Informing Patients, Activity and Motivation.
A prototype web application based on these research findings was then developed; Activity App. Activity App builds on the capabilities of an off the shelf activity-monitoring/pedometer device, FitBit, to create a service that encourages bone marrow transplant patients to remain physically active before, during and after their treatment.
As part of the development of the prototype, I worked with Sloan-Kettering to write a section of a protocol that will allow Activity App to be tested with patients as part of a forthcoming clinical trial.
While at ITP I've been fortunate enough to work on a number of health-related projects and I wanted to continue this work with my final thesis project.
In 2011 I approached Memorial Sloan-Kettering Cancer Center (MSKCC) in New York to see if we could collaborate on a project. MSKCC is one of the world's leading cancer hospitals. In 2011 MSKCC admitted over 23,000 patients, carried out 11,052 inpatient surgeries, carried out 8,181 outpatient surgeries, and conducted 1.18 million outpatient visits. The hospital also has a Design Innovation Group who work to improve procedures and patients' experiences.
Over a number of discussions we saw that there was an opportunity for me to work on a project related to bone marrow transplantation; a procedure that is currently actively being investigated and improved at the hospital. We identified the period of rehabilitation following a bone marrow transplant as being of particular interest.
We agreed that I would carry out a research and design project to improve patients' experience of bone marrow rehabilitation.
I would carry out a period research to investigate bone marrow transplantation as it's experienced today. I would then use the insights gained from this research to identify opportunities to improve the patient experience. In the time remaining I would take one of these opportunities and develop a prototype of some kind; a new service, device or application of some kind. I would take all of this work and present it the hospital for them to take forward.
In January 2012 I started working on the project.
The research phase of the project lasted just under 6 weeks. I spent two weeks carrying out desk research; reading about bone marrow transplantation and related topics. I spent 4 weeks working on-site at Memorial Sloan-Kettering Cancer Center conducting primary research. I carried out a number of in-hospital observations, visiting the locations in the hospital where bone marrow transplantation is carried out and observing staff at work.
I spent the majority of my time at the hospital meeting and interviewing medical staff, administrators and patients. I interviewed 12 people in total; each interview followed a semi-structured format and was recorded, then transcribed. These interviews allowed me to develop a detailed understanding of the journey that patients go on when they undergo a bone marrow transplant - the people they interact with, the places they visit, the objects they interact with, the activities they carry out, the knowledge and skills they are expected to acquire and carry out.
I organized the findings of this research under a number of headings:
The Patient's Journey: The experience of undergoing a bone marrow transplant is long and complicated. There are a number of different types of transplant offered depending on the patient's diagnosis, and while each type has its own set of attributes - no two transplants are identical.
Symptoms: The symptoms experienced by bone marrow transplantation patients can be debilitating. The chemotherapy regimen causes patients to feel very weak and nauseous and these symptoms can prevent patients from carrying out simple daily tasks - standing, walking. The medicines being given to patients can be harmful to organs if the patient is not properly hydrated. Patients may experience impairments to their cognitive functions; attention and short term memory. After transplantation, patients may experience a range of complications e.g. infections, graft versus host disease.
The Whole Patient: Unsurprisingly, each person who needs to undergo a bone marrow transplant is a unique individual with specific personality traits, needs, fears, expectation, hobbies and relationships. All of these factors play a role in how the person will cope with treatment and rehabilitation.
The Importance of the Caregiver: Other than the patient and the primary care physician, no other person plays a more significant role in the progress of the patient than their designated “caregiver”. The caregiver is typically a spouse, family-member or friend who remains on-hand throughout the patient's treatment and rehabilitation - this is a significant undertaking, especially when the caregiver needs to work or has other responsibilities. It would be fair to say that the caregiver is, in effect, co-opted onto the patient's medical team.
Informing Patients: Patients and caregivers receive/experience this large quantity of new information at a time when they are undergoing significant stress, and when the patient could be experiencing debilitating side affects related to their treatment. This information is not evenly distributed throughout the treatment journey, and they will be asked to make significant decisions based on this volume of new information.
Activity & Motivation: Physical activity plays a clinically important role in the patient's progress. The reasons why patients should remain physically active change over time. Before the patient receives their chemotherapy and transplant they are advised to take exercise to condition themselves e.g. walking to build up strength and fitness before being admitted. Once they enter the hospital, it's likely that they will spend some amount of time in confinement, and it's also likely that their symptoms will make physical activity difficult or unappealing. All patients are advised to remain physically active to minimize the extent to which they de-condition. After the patient leaves the hospital or confinement, it is important that they remain physically active to regain condition lost while in hospital/confinement. The more physically able the patient is, the more they are able to carry out day to day activities crucial to their wellbeing and survival. The less the patient is able to do, the less they will do - and their physical condition will deteriorate.
Once I had completed my research I investigated two design opportunities in more detail.
Opportunity 1 - Informing the Patient: I saw that there was an opportunity to re-design how useful, decision-critical information flows between the medical team, the patient and the caregiver. Rather than thinking about educating the patient using sets of materials, I proposed a way in which information could be offered through a service; experienced rather than given, staged over time, present when needed, discrete when not needed.
Opportunity 2 - Activity & Motivation: I saw that there was an opportunity to encourage patients to remain physically active throughout their entire journey. I proposed a way in which the hospital could begin to encourage physical activity long before the patient undergoes treatment. This approach/service would follow the patient during their treatment and rehabilitation; persuading them to take small amounts of regular exercise. It would give medical staff simple tools to offer feedback and encouragement to the patient; helping to keep them motivated as they return to health.
The second half of the project was devoted to responding to one of these opportunities. I chose to develop a prototype of a service that could encourage patients to remain physically active before, during and after their transplant. I called this activity and motivation service Activity App.
This pedometer is paired with a web application, Activity App. Activity App offers a simple view of the patient's level of physical activity without distractions and inappropriate game mechanics, options and features.
Each day, patients respond to their reminder SMS rating how difficult they found yesterday's activity. This gives medical staff vital, subjective information about the patients ability to remain active.
For the first time, rather than recommending physical activity to patients at different times and in different ways - MSKCC can offer a simple service that follows a patient throughout their treatment.
The service allows patients and therapists to work together to set a goal, and rather than expecting the patient to be active, the service expects them to be inactive - and offers a “hot trigger” in the form of a daily SMS that encourages them to be active.
Patients can monitor their own progress and can offer simple, subjective feedback. Because patients find feedback from medical staff highly motivating, medical staff have some simple tools they can use to offer encouragement.
Activity App is a fully working prototype, written in Sinatra and deployed to Heroku. I have released the source code to Memorial Sloan-Kettering Cancer Center under an MIT License for them to continue its development.
I presented my thesis project to the ITP community and representatives of MSKCC's Design Innovation Group in May of 2012 - completing the requirements for my Master's degree.
During the project I was asked to write a section of a clinical trial protocol that would allow Activity App to be used in a forthcoming clinical trial involving a particular population of bone marrow transplant patients. and so if the hospital are happy that Activity App is stable and secure enough to be used in a real world context, it can be. This is a really exciting prospect for me.
I have completed a full write-up of the project, including a set of recommendations about how Activity App could be developed further, and delivered this write-up to Memorial Sloan-Kettering Cancer Center. I would love to continue to collaborate them in some way.
I am very happy with the work I did in such a short period of time. It would now be desirable for Activity App to be tested with real patients and medical staff. During this test a small group of patients would be offered the service at the beginning of their treatment. They would use the service, and offer feedback at set intervals. This feedback could be implemented as it was given by patients, or it could be aggregated and addressed at the end of the test. I have worked with Memorial Sloan-Kettering to write a section of a clinical trial document that would allow Activity App to be tested in such a way, and I hope that this happens in the near future.
I would also be very keen to see this type of service rolled out to all patients at the hospital. It may become standard practice for medical staff to recommend a period of conditioning/activity to all patients in advance of their treatment. If this was the case, then offering a service like Activity App could help to support patients’ motivation while offering medical staff insight into patient behavior. Over time, the ways in which patients are supported in the use of these types of persuasive systems could be refined and tested further.
It’s my belief that the integration of physical activity in a medical context, supported by simple technologies like Activity App, can improve clinical outcomes – obviously this is something that should be carefully considered, introduced slowly and evaluated over time. It’s an exciting challenge.
The key thing I learned from this project was that I am highly motivated and produce good work when I can work reasonably independently on projects that are personally meaningful to me. My challenge now is to make or shape an environment that allows me to work in this way most or all of the time; it’s likely that I will start or help co-found some sort of company/venture that takes on difficult challanges at the intersection of humanity and technology and seeks to measurably improve people’s lives.
While the project was incredibly motivating, I must also recognize that working independently on a challenging project, especially one to human health, can be lonely and stressful at times. I had no contact with upsetting situations while carrying out research at MSKCC, but I underestimated how much I would be affected by the topic itself, the language, the enormity of the challenge facing medical staff and patients. This context, combined with a very challenging schedule in the early phases of the project, weighed on me. If and when I do a project related to human health in the future I will be sure to consider the impact of the subject matter on my emotional well being and mental health and I will put measures in place to ensure that I have proper support.
I am satisfied that the methods I selected to undertake this project were, for the most part, appropriate and useful. The decision to adhere to a reasonably detailed project structure allowed me to spend the majority of my time researching, learning, interacting with people, documenting my work and producing software. I spent relatively little time during the project on planning or task management.
While each phase of the project was completed on time, or within an acceptable margin, I see now that this was in-part due to my willingness to scale the amount of work I did in each phase. I could have spent more time in each phase – interviewing more people, adding more functionality – but chose not to. I had never really reflected properly on the relationship between the attitude of a project (doing enough things well, rather than doing as many things as possible) and its outcome. I take heart in this and will continue to develop my skills and understanding of project planning and pre-production.
Conducting a period of secondary research was very useful. In a short period of time I became familiar with the basic facts, processes and vocabularly associated with bone marrow transplantation. Without this knowledge I wouldn’t have been able to ask meaningful questions when it came time to do interviews.
In retrospect I feel as though my secondary research could have been improved; all the materials I read didn’t fit closely enough with my research questions, and I could have made better use of my time if they had – I would have been able to read more deeply in a smaller number of areas. If I was to do this project again I would establish a much closer link between my research questions, a set of topics related to those questions and a set of materials related to each topic. I would consult with experts who could recommend materials to me. I would work with a reference librarian to help me find materials closely aligned with my research questions, and I would read deeply on those topics. I need to improve my secondary research skills.
The observation exercises were useful and produced useful photographs and notes. While carrying out the observations I felt uncomfortable and out of place and my interactions with people felt forced and shallow – I need to practice these skills so that I am more used to the being an outsider, being in the moment while gathering useful information.
Carrying out semi-structured interviews resulted in a wealth of detailed, useful information and allowed me to get to know the people who are involved in the delivery of care. I drew heavily on the information gathered during these interviews later in the project, and the empathy I developed for the people at MSKCC helped me make good decisions when it came to developing my ideas and software.
While I’ve carried out research interviews many times in the past, I found these interviews to be very challenging – the subject matter was complex and there was no real opportunity for people to be re-interviewed; it was important that each interview went well. If I was to do the interviews again I think I would reduce the number of questions significantly and tailor each interview more closely to the interviewee – I could have asked fewer, better questions and allowed the conversation to flow more naturally; in a sense I need to think of interviews as structured conversations rather than interviews where the person has information and my questions are tools to prompt them to give me that information.
Also, the volume of information created during interviews became something of a problem in itself. From a purely practical point of view, transcribing interviews is tedious. I could have used a commercial service to do the transcriptions, but felt as though there was merit in “sitting-with” and re-listening to interviews; transcribing them myself.
I am now sure that, given the short amount of time available in the research phase, having the interviews transcribed by sombody would have been a better approach. I would have had more time to do other work, and once the interviews had been transcribed I would have had more time to reflect on the things people were saying. I will take this approach on my next project.
The prototyping and production phases of the project ran reasonably smoothly. Once I had described the practical project I wanted to work on I went a level deeper to describe the things I wanted users to be able to do. I used post-it notes to break “stories” (e.g. the user logs in for the first time) down into individual steps and supporting functions (e.g. the user must be able to pair their fitbit with my software). This system was simple and effective and one I’ll use again.
The production of the software itself was highly iterative; I released code to the production server over 140 times in the 4 week prototyping phase. I was able to do this because I was using the domain specific language Sinatra and the hosting provider Heroku. This system allowed me to write code on my laptop and issue a single command to release code to the production server. Using live data from my FitBit and testing with two volunteers allowed me to test that each of these small releases worked and improved the user experience. I’ll be sure to bring this approach forward with me; when I develop software it really matters to me that I’m using real data. It’s important that my tools help me, rather than getting in the way – I am just not interested in configuring servers, so a system like Heroku is perfect for me. Being able to work locally and push to a remote server means that I’m able to see my changes almost instantly. This process was really groundbreaking for me.
I am glad that I gave myself a specific window of time in the project to document and present my work. I didn’t, however, leave myself enough time to document my work as I went along. I should have had one day per week, or one week in the middle of the project when I drew togehter all my materials into some sort of collection. Doing this would have made my final documentation more of a compilation exercise, rather than a separate act of documentation. This is something I’m still thinking about – but on my next project I’ll consider whether more “real-time” documentation would be useful.
I owe a sincere debt of gratitude to the staff of Memorial Sloan-Kettering's Design Innovation team for their support and guidance throughout this project. I would particularly like to thank Era Bani who acted as my liaison at the hospital and made sure that I had access to people, facilities and information vital to my project.
I am incredibly grateful to patients, doctors, nurses, therapists and administrators at Sloan-Kettering who, though busy with far more serious matters, agreed to be interviewed and provide feedback during the course of the project.
I would like to thank all the staff and students at ITP for their support. I particularly want to thank Nancy Hechinger for her trust, patience and straight-talk. Steven Dean made the initial introduction to MSKCC and instructed me in the design of health-related applications. Steven Klise offered technical help and sound advice while I was developing the prototype of Activity App.