Adopting the HARMS Program

Adopting the HARMS Program

Implement the HARMS Program into your clinic

If the HARMS Program interests you, then read on about how you might consider adopting it at your clinic.

Clinic Approval: The first step is to get clinic approval. The education part is really important here. If you don't get clinic-wide approval then it can still be done for just a few prescriber's practices, but it creates some tension in that some patients at the clinic are subjected to UDT because their doctor is Dr. X, whereas patients of Dr. Y don't have to do them.

Education: Education is critical as well. We are creating resources to help with this. Education is focused around three different roles within the HARMS Program:

  • Clinic admin staff
  • Physicians/prescribers
  • Patients

Creating a list, checking it twice...: In the beginning, you will need to make a master list of every patient that should be in HARMS, and what the risk category is. This will allow clinical administration to know who to test and how often. What we did at MFHT is we pulled the data for all patients that were prescribing an opioid in the last 12 months, and then sent the list to the prescribing physician. Patients were then removed from the list if it was a short-term prescription, they had a cancer/palliative diagnosis, or other concerns (nursing home where medication was supervised anyway, severe mobility issues, etc). Risk category was assigned, and the information was sent to a central admin person who kept the list.

If a new patient is started on opioids for CNCP, then the name and risk category should be sent to the central clinical admin at your clinic. This will ensure that that person is on the master list for UDT in accordance with his/her risk category.

If a patient has opioid medication discontinued or moves, it is important to remove patient from the list.

When deciding who to include on the list, we decided it was easiest to use "someone who is on average, using one opioid tab/patch or more per day". While it may seem better to use morphine equivalents per day (MED), this requires more effort and we wanted to keep things simple. So if someone is only getting 10 Percocet per month, he would not be in our HARMS Program. However, if someone is on 30 T#3 per month, then he would be in the program.

Maintenance of EMR Chart: At our clinic, the keeper of the master list writes down in the same section of the chart the risk category and when he/she started the program. We have OSCAR and use the "reminders" section so we all know where to look. This is just added assurance when we refill an opioid that patient is on the master list.

Randomization of UDT: There are numerous different ways of doing this. We initially just randomly selected 10% of patients in the low-risk group per month. While this may be an easy way to randomize, the downside to this strategy is that mathematically about 28% of people won't be randomized at all during a given 12-month period. The upgraded HARMS program instead will assign a certain number of UDT over the course of the year (1 for low risk, 2 for intermediate risk, 6 for high risk) and then during that month patient will be called to provide a sample. In addition to those guaranteed UDT, patient may be randomly selected for additional UDT (only about 5% per month). This guarantees everyone will be selected in a frequency accordant with risk (and at least once per year). The small randomization component is simply so that patients will never be able to predict if they will or won't get called for a random. Otherwise, if a low-risk patient provided one UDT and he knew he was only providing one, then he could do whatever he wanted knowing that he wouldn't get selected again until the next 12-month period.

Conduction of UDT: There will be a list of patients for a given month. They are randomly distributed over the course of the UDT clinic days that month and notified of the apt ~24-36h beforehand. If patient is unavailable, no-shows, cancels, can't urinate, etc then he/she goes onto a recall list so that more attempts will be made to collect a UDT. If there are ongoing concerns and patient is not providing a UDT then the prescriber is notified.

Once patient presents to provide a sample, our clinic follows a protocol (will post here soon). We use the START-IT tool to collect and interpret the IA component of the UDT. Currently, we send all of our UDT results for LC-MS. We are looking at using START-IT to inform the health care assistant which UDT shoudl be sent, and which shouldn't, to save the system costs from these tests.

Communicating Results of UDT to physician/prescriber: The results are then inputted into the EMR with the click of a few buttons (assuming you have Acuro, OSCAR, or Practice Solutions). The physician, depending on his/her preference and those preferences of the clinic, can either be notified of all results or just unexpected results. Regardless, the LC-MS result will come to the physician. If the physician wants to adjust risk category based on that result, he/she can simply message the central clinical admin person to do this.

Obstacles: While the HARMS Program is practical and simple, there are a few potential obstacles to implementing it. Here's a list we came up with, along with some suggestions:

Patient reluctance:

  • the MFHT encountered some of this, but because it was universally implemented at the entire clinic, no patient could reasonably say that we were picking on him/her. The requirements for patients that are low risk are minimal - essentially one UDT per year. Consider patient education either before or during the roll-out. This could include posters around the clinic, or public information sessions as we did.

Clinician reluctance:

  • there could be push-back from prescribers that this will create a lot of work. It won't. In fact, it has saved clinicians a lot of worry and stress by identifying high risk patients and guiding subsequent management. All that the clinician has to do is exclude patients if needed (palliative, cancer, etc) decide on a risk category, and then act on the result. The coordination of UDT and performing it is all done by clinical admin. In Ontario, most physicians can bill OHIP for UDT as well to compensate for the time to interpret them.
  • there could also be push-back from clinicians that we all have different practice styles and we shouldn't conform to a program. If you actually read and understand HARMS, it's not meant to dictate what a clinician does. It's meant to give guidance in a systematic way, which when incorporated with clinical Gestalt of the patient, and differing practice style of the prescriber, could lead to very different actions based on the same UDT result.

Human resources:

  • "we don't have the resources to conduct UDT at our clinic, so HARMS isn't possible". Not true, there are numerous options here. Firstly, consider signing up for our pilot study. We will provide funding to cover human resource costs, IA panels, etc and faciliate conduction of UDT at your clinic. The IA conduction can generate it's own revenue stream (although it's billed by the clinician). At the MFHT, we have ~4000 patients, ~100 on opioids for CNCP, and we initially used our healthcare assistant 2 half-days per week to complete all of the UDT needed (that time was even enough to conduct UDT ~bi-monthly on ~30 patients we have on opioid agonist treatment). If you're really stuck, you could conduct UDT at a lab but there are numerous limitations to this (no self-report, no START-IT for automatic interpretation, and no assurance that the urine provided is actually the patient's as labs have been known to give patient the specimen container to go home and then bring back the sample - very open to tampering).


  • if you do the UDT in your clinic you need a bathroom. See our protocols section for how we conduct UDT. We book two half-days per week to do UDT so that we don't have to constantly be getting the bathroom ready with blue water.

Treatment agreements:

  • MFHT had every patient sign a treatment agreement. This was, honestly, a hassle. Keepign track of it, explaining it to everyone, and then documenting/logging them in the EMR. You could reasonably consider not doing treatment agreements but instead displaying posters around the office telling everything that the HARMS Program is being adopted and that compliance is an expectation, with a link to our website or other educational material. There are a few alternatives if you really want a treatment agreement - with the next version of START-IT we can include a treatment agreement in there that patients can electronically sign, to at least save some of the paperwork hassle. You could also just have patients sign the agreements on paper the way we did, as they come into the office. It only takes a few minutes - especially if you give the patient the agreement to read at home, and tell him/her to sign it and drop it off when completed.