1. People forget
One of the participants forgot their medication on the very first day of the trial. Not a very good start, but it might be representative of behaviour when taking medication. However, because most people who are willing to take medication at least take the first one in an effort to reduce their symptoms I had to find a way of getting them started, at least for the first pill.
The solution
Remind all participants (in both groups) to take their first pill. For the non-intervention group, this will be the only reminder they will receive during the study.
2. The equipment needs to be thoroughly tested
MessageBird allows you to set up templates for later use. It was one of the reasons I chose to go with them as the provider for sending SMSes. In order to fully utilise this feature, I spent a lot of time designing and creating an automated workflow.
Working on a mac I noticed that the webapp wasn’t always being responsive. Sometimes I’d press the “save” button once and it would work, sometimes the site just hung until I gave up and refreshed the page to try again. For the pilot study, I had to manually send the SMS reminder at the appropriate time. This isn’t ideal of course as when I run the full study I won’t have the luxury of being able to watch the clock and send out each SMS at the designated time.
Also, don’t use the “quick send” option, as this does not pick up the personalisation
The solution
I used a PC to finish creating the templates. The website worked without issue on a desktop PC. I will continue to test the automation of the messages throughout the pilot study to ensure that the messages are always delivered.
3. People question the placebo
Both pilot study participants questioned why the packet of tic-tac had been opened and why it was half full.
Solution
Inform the participants when introducing the study, that I will be providing them with a weeks supply of placebo medication.
Also, explain to them why I have chosen Tic-Tacs (small and no-one hates Tic-Tacs) for the placebo.
4. The timing needs to be just right (or as close to it as I can manage)
When starting the experiment I created the morning alert to be at 8:15am the time the participant was getting ready to leave for work. In hindsight, this was a little too late to remind them. When discussing their experience and the reasons why they missed pills, I was informed that;
If I’ve got a text message when I’m running around trying to get out the house I might not see it or ignore it.
Solution
Designing a way of prompting the user to put the pills in their ‘everyday carry’ so that they remember to take the pills with them if they need to take one them at work. Or, finding a way to alert the user after they wake up, but not close to the time when they need to leave the house so that they aren’t in a rush and will be more prone to forgetting, or ignoring the prompt.
5. Routine’s change
One of my participants works a permanent Monday to Friday 9-5 job. On the weekend, he did not pay much attention to his phone; often leaving it around the house instead of keeping it in his pocket as he would do when going to work during the week.
He also woke up much later on the weekend, so the 8am alert was completely missed. Because of the change in routine, he also did not notice the 6pm alert.
Solution
Allow participants to specify when they want to be alerted on the weekend (if they work 9-5 jobs) or more broadly ask them to specify if there are any days that they should be alerted at a different time; providing they can still take their medication within a 2-hour window.
6. Not everyone has notifications turned on
Luckily enough, I was with one of the participants when the SMS message was received. They had guessed it was their pill reminder when they heard their phone SMS sound because they had gotten used to the timing of the delivery. I asked to have a look at their phone to see how the notification appeared and was surprised to see that nothing showed up on their home screen. On probing, I found out that they had turned off their notifications long ago because “it annoyed them”
Solution
Ask that all participants have SMS notifications turned on during the period of the study.
7. People over-estimate their adherence
When asking one of the participants how many Tic-Tac’s they missed they stated two, when in fact, they had missed four. I knew this by doing a pill count of the remaining Tic-Tacs in their box.