Chapter 4: UDT Selection

Now that we have all of our HARMS Program patients risk-stratified and on a master list, the next step is actual selection for UDT…

While HARMS was designed to apply UDT in a randomized, risk-concordant fashion, there are times when it is appropriate to do “non-randomized” UDT. Namely, if a physician has clinical concerns for whatever reason (early refills, deteriorating social function, etc.) then he/she may request a UDT. Therefore, two types of UDT can be ordered – randomized (through HARMS) and non-randomized (typically clinician concern or baseline UDT for a new patient).

Randomized: this is the core of the HARMS Program UDT system. We use computer randomization to select a patient for a certain number of months per year to provide a UDT. The number of UDT is based on risk category, with low risk patients getting selected for one month per year, medium risk patients two months, and high-risk patients six months. This is based partly on the following expert recommendations.1

  • At least annually for low risk patients
  • ≥ 2 times per year for moderate risk patients
  • ≥ 3 per year for high-risk patients
  • Additional monitoring at any level per clinical judgment1

Given the significant gap between the recommendation for medium risk (2 times/year), and our “structured” stream (2-4 times/month), we felt as though the high-risk patients should be subjected to 6 times/year so we chose this instead of “≥3”.

We also wanted the capacity for additional randomization so that, in theory, anyone in the program could be selected any month. This would mean that anyone could be selected at any time, whereas if we didn’t do this then a low-risk patient selected in the first month of the year would know that there are no more UDTs for the next 11 months. We therefore added an option to randomly select a small number of patients (5%) each month, in addition to the 1/2/6 months per year already randomized based on risk category. This is also built into the randomization spreadsheet and is found on the UDT Selection List tab.

Non-randomized: these UDT are typically fairly straight-forward. If a physician requests a UDT (due to an observational concern, baseline UDT for a new patient, etc.) then a UDT may be requested independent of the above randomization algorithm. At our clinic, we do this by sending a message to administration to book a patient for UDT. The patient can then be manually added on to the list for the coming month. The other time where a patient may provide a non-randomized UDT would be if he/she is in either the structured risk category (Level 5) or opioid addiction (Level 6), in which case he may be providing UDT on a scheduled basis (i.e. every 1-2 weeks, monthly, etc. depending on individual physician preferences). With these structured UDT, we have found it helpful to put the onus on the patient with clear instructions that he/she is responsible for booking UDT and attending them at regular intervals.

In terms of actually generating the randomization list, we would strongly recommend using the spreadsheet tool that we created and currently use in our clinic. That document has explanations about how to generate patient lists for booking staff each m o nth that ensure randomization frequency is concordant with risk category, and to simplify the whole process as much as possible.

Note that there is significant variability in how you may go about informing/consenting patients to be a part of your HARMS Program. At our clinic, we rolled the program out slowly and had a discussion with each patient as part of a routine medical visit. As part of that discussion, we completed a treatment agreement that included consents around UDT and the HARMS Program. See Appendix I and IX for more information.

Chapter Pearls

  • While HARMS was built to support randomized selection for UDT, if a concern exists then the physician can (and should) still message clinical administration to book a "non-random" UDT.
  • Consider using the spreadsheet that we created - and use currently in our clinic - to both maintain a master list and generate patient lists to provide random UDT.


  1. Argoff CE, Alford DP, Fudin J, et al. Rational Urine Drug Monitoring in Patients Receiving Opioids for Chronic Pain: Consensus Recommendations. Pain Med Malden Mass. 2018;19(1):97-117. doi:10.1093/pm/pnx285

Now that names have been selected for UDT, we will go over how the booking staff notifies patients of the UDT appointment and what to do in the various situations that may arise…