Real-life observational studies are essential to evidence-based practice change. Sure, they're not the gold standard for proving the efficacy of a device or treatment (that's what randomised control trials are for), but they have the ability to capture a patient sample that represents the average population, and can prove whether patients in routine practice are achieving expected outcomes [1]. Conducting observational studies in a clinical setting however, is not always easy, with time constraints, financial demands and a lack of supportive research infrastructure presenting common challenges [2].
Having worked with a number of clinical partners with various levels of clinical and research experience, it got us thinking- how exactly do resource and time limitations affect observational studies in clinical practice? And what advice can we provide to those wanting to establish their own investigations? With over 11,000 hours of activity clocked by our team in 2018, we dug deep into our project management archives for some answers.
"We're good to go!"
We've heard that phrase one too many times. Variations include: "I've got all the data", "We've got ethics covered", "We've got all our patients", "I've got a fellow starting next month- he can do it". That's all great, but what variables does the literature say are worth collecting? Does your ethics allow you to retrospectively add patients to your study? Do you have the right sample size to answer the question you're asking? How many hours is your fellow planning on committing to research each week?
If there's one thing we place a lot of emphasis on, it has to be this- everyone needs a research plan. Whether it's a grand plan encompassing a whole patient registry and cohort studies, or a single research protocol for a simple investigation, you need a thoroughly-explored, well-documented plan in place before commencing. Why? Because in our experience, we've found that studies that took off without an adequate plan in place took up to 4 times longer than expected. That's 2.5 years for a study that you thought would only take 8 months. And with that comes a whole array of problems- going over budget, dealing with staff turnover, and in some cases, losing the novelty of your idea because the literature is now flooded with similar reports.
Having said that, even carefully planned studies can encounter a number of problems that affect both timeline and budget, and there's two common reasons why:
Scope creep
There's a reason why they refer to it as "creep". It's because you think it's within the scope of what you've planned, but you don't realise that recruiting a few extra patients means days of extra follow-ups, or doing image analysis means spending hours on the phone with customer support in a different timezone. Sometime it's outside your control- the hospital decides to change their MRI scanner halfway through the study and you're stuck deciding between re-scanning your patients or abandoning the project all together (true story). With the work we've been involved in, we found that projects that went 3-4 times over budget did so mainly because of scope creep, with over-committed staff another key factor (more on that later).
The best way to minimise scope creep is to plan, and to plan adequately. And if you're trialing new methods, we highly recommend conducting a pilot study on a sample subset. This gives you a better feel for what to expect in terms of time and cost, and enables you mitigate risks and handle roadblocks that might come your way.
In-house research
Conducting research in-house via practice staff, students or resident medical officers definitely appears to be the easier and cheaper option, however, our experience indicates otherwise. Some carefully planned studies we've been involved in came under budget by up to 50%, but still took 2-3 times longer than expected. And the key reason for this is that there is a significant disconnect between what clinicians assume their in-house staff are willing or able to do, and what the staff can or actually deliver.
Sometimes the staff are spread thin across multiple projects and other administrative duties. Other times, research is low on their priorities list because they need to keep the practice running smoothly. For some resident medical officers, it's just a box they need ticked and a new person will clean up after them in next year's rotation. Medical students are usually keener, but less experienced and more likely to require close supervision and guidance. While it seems cheaper to get staff already on the payroll to do research, they are costing you indirectly in terms of time spent training and monitoring, and time taken away from other duties.
Tempering expectations
So what has our experience told us are the key considerations for conducting successful research in clinical practice?
The first thing is (you guessed it!) to put a research or study plan in place. Factor in contingencies; these include delays to the ethics approval process, higher than expected patient drop-out rates and patients lost to follow-up. Pilot test your methods- figure out how long it takes to retrieve data from paper and electronic records, and extrapolate that across your expected number of patients.
The second and most important thing is to set realistic expectations. Research in clinical practice is often oversimplified and under resourced. Know your budget, check the literature, and be realistic about clinical flow, methods practicality, and staff or resource commitment. Don't ignore or underestimate any training, coordination or logistics that need to be accounted for.
Temper your expectations, and factor them into your plan of action. An achievable plan is far better than an overly optimistic one. Conducting research is an investment of your time and resources, with evidence-based practice change a return on that investment. The key to successfully reaching that end state is to be real about what can and cannot be achieved within the constraints of clinical practice.
But what happens when you don't have time to adequately put plans in place, or the experience to scope all the things that could go wrong? Luckily, outsourcing to a team that manages and delivers research on the daily is an option. With 11,000+ hours of research experience in the past year alone, we're keen on hearing about your ideas, and helping you put them into action!
References
Cohen A.T. et al., European Heart Journal Supplements, Volume 17 , Issue suppl_D, July 2015, Pages D2–D8
Institute of Medicine (US) Forum on Drug Discovery, Development, and Translation. National Academies Press (US) ; 2010. Challenges in Clinical Research. Available from: https://www.ncbi.nlm.nih.gov/books/NBK50888/