When you're conducting clinical research studies in resource limited settings, and multinational settings, what we might loosely think of as global health research and international multi-center clinical trials Then you may face a new set of challenges in all aspects of clinical research data management. We've already discussed some of those in the context of multi-center studies. Now we're going to take a quick look at some additional challenges. But if you are currently working in such settings. I encourage you to start a discussion on the forums to share your experiences and explore the topic further with other students. In this course so far, we have considered designing your clinical research study, identifying best approaches to capturing your study data, paper verses electronic systems for example. And designing your data collection forms. Now we're going to rethink those steps in the context of resource-limited settings, which often means low income countries, middle income countries also. Though many of these concepts also apply to work in rural parts of the United States, for example. It's important to understand your setting, rethink your data collection plans, and redesign your data forms as necessary. Our focus will be on multi-interclinical research studies where you need to be in touch with coordinating center. But these general lessons can apply to single site studies too. To start, there's an ongoing debate about whether it's ethical to conduct clinical studies in developing countries in the first place. Can researchers, in these settings, effectively explain the motivations of research? And can patients truly give conformed consent? Or are research institutions and drug companies exploiting populations in order to save 60% of the expense of a study? Even in collaborative studies, it's important to ensure one site isn't providing all the data while the other gets all the scientific credit. It's an important debate but not the focus of this course. I'd recommend you take a look at this great January 2013 summary by debating matters, which has arguments for and against with plenty of relevant links to media reports and the scientific literature. To get back to data management, your solutions for a clinical research study may need to be designed to accommodate limited infrastructure and resources. Here are a few of the key limitations that may affect data management capacity specifically. Many of the approaches we've discussed in this course rely on computers, computer networks and mobile devices to handle clinical research data. That is to collect it initially, curate it and monitor its quality and to analyze it. Well, those computers run on electricity. The availability of stable sources of electricity is important to most clinical research functions, to run your equipment and charge batteries. Frequent blackouts, brownouts or a lack of electrical service entirely will change your data management plans. Voltage fluctuations too can damage sensitive electronic equipment. The availability of Internet access should also be a concern when developing your study plans. If you are conducting a multi-center, multi-national trial with performance sites that have no Internet access, sporadic Internet access or slow Internet access then web based data entry like we have been using like with red cap in this course isn't realistic. Its likely that activity is like hands on teaching monitoring and data auditing will need to be done in person. Conference calls may rely on mobile phone connectivity, instead of having a group meeting with web conferencing software. And even if the study site has fantastic internet service, there may be an unacceptable, unacceptable network lag. If you're data entry system is hosted in the US and your study is in Australia. Test all of these considerations before you deploy your data capture solution. For the data equipment that you do have, will you have the necessary environments? Will you have a secure locked room, even monitored if necessary, where your documents or equipment with PHI can be stored? High powered servers may need climate controlled environments to function. But so did paper documents for long term storage. Even regular devices, tablets, laptops, failed to work in very hot, cold or humid conditions. Have you ever seen one of those devices too hot to operate messages on the smart phone? It's likely you also rely on a shipping service for deliveries of steady equipment. Study drugs, lab equipment, data capture forms, etcetera. Can you receive these in a timely fashion? Will you have reliable phone service to join meetings, report serious adverse events, or request support from your coordinating center. All of these services may be available and reliable to a degree that fits your study. Or they may not. These are great conditions to include in your handbook of operating procedures. For each different research setting, you will likely have different regulations with which you need to comply. National guidelines may determine whether you can store blood samples or ask about race in a survey. We generally covered US regulations for data privacy and security in this course. But you may need to comply with stricter requirements in other countries. Many research groups in developing countries face challenges hiring and retaining data management personnel. General computer literacy may be low. And the training and the certification programs often aren't available. So anyone with these skills is in high demand. Research teams can benefit from structured training plans to develop the necessary skills in newly hired staff. Along with lots of supervised hands-on practice. If you have study sites in diverse settings, it's likely you'll need case report forms in multiple languages. These'll need to be translated and back-translated, tested in context, and reviewed by the appropriate IRVs. Test the meaning and interpretation of your form and survey questions with local personnel too. Even if you're familiar with the language. And if you're a coordinating center dealing with study sites across multiple time zones, be prepared for additional complexity in tracking events and providing support and training to your sties. The attitude of people in your study setting towards research plays a huge role in both your study and your data plan. I've encountered researchers in extremely resource-limited settings whose support team shares a passion for health research. I've also spoken with researchers who did not have the support of their hospitals or institutions. One researcher had to bribe her hospital's record room staff in order to access records for research. Be a positive phase of clinical research to establish the necessary networks of trust. There are also equipment challenges to consider. If you're planning a computer-driven data collection in resource-limited settings, how are you going to get the equipment and necessary software? That can be expensive to order and shipped to you. There are often import restrictions or hand carrying them as a traveler restrictions on how many laptops you can carry back into a country. We've already talked about electricity, but you may also need everything from battery backups and surge protectors to server racks and generators. You should also document in your SOPs how you will handle maintenance and repair of the equipment. The estimated turnaround time will help you plan how many replacement items you should have on hand for immediate use. Not all software may be available in the country you're working in or it might not be available in the local language. I worked with one clinic in a middle income country. That indicated they didn't want to use tablet computers for data collection at the point of care. Because it gave the appearance of wealth. And they felt that delivered the wrong message to the underprivileged population they served. In other places carrying expensive equipment can make you a target. You would be risking not just your technology and data, but your personal safety. It's not worth it. On the other hand, mobile flip phones seem ubiquitous in some research limited settings. So they might present a good data collection tool. Designing data collection interfaces for mobile devices would be a good course topic in itself. And if I've haven't included it in the first course iteration because of time considerations, we acknowledge its importance. Paper can be an excellent solution when working in resource-limited settings for all the reasons described on the previous slide. Just don't forget, paper forms are more error-prone and require more work to prepare data for analysis or to send data to a coordinating center. Many of the resources you might use for retrospective view are on paper. Medication dispensing books, lab results records,patient records but electronic health records, laboratory systems and electronic pharmacy systems are becoming more prevalent. Open source solutions like Open MRS are very popular. And I've seen many locally developed customs systems also. Finally let's take a look at some examples of how understanding the local setting may change the design of your form questions. There aren't great available guidelines on this so I'll share with you some examples of challenges I've learned from researchers working around the world In one data collection effort in the Caribbean, local researchers explain that a participants full name and birthdate are ineffective as reliable identifiers. Sometimes the study participants never knew their exact date of birth, and high illiteracy meant that many could not spell their names. Different phonetic spellings by different study personnel, led to confusion on forms. So they began to collect mother's name, as an extra identifier in these populations to help distinguish individuals. It was a new and unexpected field to include on form. Similarly, frequently changing address information for transitory populations. Changing or nontradiotioal means of contact etcetera may require forms that queery these items more frequently. I heard another example from a local researcher in a Southeast Asian country who was responsible for distributing surveys to a number of health care clinics. One question on the survey read, please indicate which support services are available for persons followed by this facility. Circle the numbers for all applicable responses. And this was followed by a long list of options. According to this investigator, personal at the sites interpreted this differently. Culturally, they considered it impolite to say no and therefore felt uncomfortable responding that a service was not available. So they often marked, yes it is available, for the major it services listed by circling most of the numbers. They would accept that available might mean after a six hour drive to the nearest city. But that wasn't the response the survey designers intended to elicit. In comparison, a different survey question allowed respondents to say if whether serv, services were available onsite, offsite at a distance, or not at all. And adding these shades of grey produced more representative answers. We've covered in a previous lecture, the utility of numeric coding. Sometimes good, sometimes excessive work. But 0 for no and 1 for yes is a very common mapping. One data manager in a French speaking country explained to me why his team didn't use that system though. They found that the 0 for no was visually similar to the o often used as an abbreviation for oui meaning yes in French. This created a sort of, cognitive dissidence with data entry personnel, and resulted in higher error rates, despite training. When they switched the coding to 2 for no, the error rate declined. Finally, pediatricians I met from Central America had no pediatric scales in their clinic. Like the scale pictured here, they only had an adult scale. So in order to measure infant weight for their internal study, they recorded the weight of the mother holding her infant, and then the mother alone, and subtracted the two. Their paper forms contain spaces for all these measures and the resulted calculated infant weight. And once they received the pediatric scale they had to redesign there forms. But in short, develop forms, but make sure you also test them in context. If you're study sites are located in divers countries, you'll need to assure that those forms are electronic data collection systems. Makes sense to local staff based on workflow and context. This video was a brief overview of the challenges of conducting data management for clinical research in resource limited settings. We've reviewed the importances of understanding the differences in the settings for your study. Rethinking your data collection modalities, paper isn't always bad. And redesigning your data collection forms as necessary.
