Background:​ Up to 40% of patients are prescribed opioids prior to total knee arthroplasty (TKA) in the USA. These patients prescribed preoperative opioids have increased complications and worse outcomes after surgery. But, the impact of preoperative opioid use on patient-reported outcome (PRO) scores after TKA has not been studied extensively. To our knowledge, the number of patients prescribed opioids before and after TKA in Canada also has never been reported. The purpose of this thesis is to estimate the prevalence of opioid use before and after TKA in Alberta, Canada, and to determine the impact of preoperative opioid use on PRO scores 12-months after TKA.

Methods:​ A systematic review evaluated the impact of preoperative opioid use on PRO pain and function scores after TKA. We then described three methods that can detect consistent opioid use with administrative health data. Applying this methodology to a cohort of patients that underwent primary, elective TKA between 2013 and 2015 in Alberta, Canada, we estimated the rate of opioid use before and after TKA, along with the dose, duration and most common opioid formulations dispensed. We also analyzed the relationship between preoperative opioid use and patient reported outcomes 12-months after TKA, adjusting for potentially confounding variables with multivariable linear regression. Potentially confounding variables included patient’s age, sex, comorbidities including depression and preoperative score.

Results:​ The systematic review consisted of six studies that all reported patients prescribed opioid prior to surgery had worse clinical outcomes after surgery. The included studies had a moderate to high risk of bias as outcomes did not adjust for potential confounding factors such as preoperative PRO score or a history of depression. All studies were published from centers in the USA.

In our retrospective, multicenter population-based study, 31% (n = 592) were prescribed opioids prior to TKA. Those patients (n = 124) that were considered long-term opioid users had worse adjusted Western Ontario and MacMaster Universities Osteoarthritis Index (WOMAC) pain and function scores 12-months after TKA compared to preoperative opioid naïve patients (pain score beta coefficient 7.7 [95% CI 4.0, 11.6], function score beta coefficient:​ 7.8 [95% CI 4.0, 11.6];​ p<0.001. Few (<1%) patients that were not dispensed opioids prior to TKA became long-term opioid users between 180 and 360-days after surgery, but 44% (n=55) of preoperative long-term opioid users remained long-term opioid users 12-months after TKA. Tramadol, codeine and oxycodone were the most commonly prescribed opioids before and after TKA.

Conclusion:​ A significant number of patients were dispensed opioids before and after TKA in Alberta, Canada and patients dispensed preoperative opioids had worse pain and functional outcomes 12-months after TKA when compared to those who were not. These results substantiate previous work that suggested patients prescribed preoperative opioids should be judiciously counselled regarding expected outcomes after TKA as they have been observed to have more complications and worse outcomes after surgery. Future research is needed to define risk factors associated with persistent postoperative opioid use, and to determine whether weaning opioids preoperatively can optimize outcomes after surgery.