@article{Xiong_2017, title={Builiding Methods for a Proactive Prescription Drug Surveillance System}, volume={9}, url={https://ojphi.org/ojs/index.php/ojphi/article/view/7591}, DOI={10.5210/ojphi.v9i1.7591}, abstractNote={<div style="left: 90px; top: 314.091px; font-size: 14.1667px; font-family: sans-serif; transform: scaleX(1.08129);" data-canvas-width="63.778333333333336">Objective</div><div style="left: 105px; top: 329.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01716);" data-canvas-width="379.14674999999994">This study aims to show the application of longitudinal statistical</div><div style="left: 90px; top: 345.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.97116);" data-canvas-width="394.9383333333335">and epidemiological methods for building a proactive prescription</div><div style="left: 90px; top: 362.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00062);" data-canvas-width="241.5841666666667">drug surveillance system for public health.</div><div style="left: 90px; top: 394.091px; font-size: 14.1667px; font-family: sans-serif; transform: scaleX(1.11768);" data-canvas-width="82.63416666666666">Introduction</div><div style="left: 105px; top: 409.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.995988);" data-canvas-width="378.30949999999996">Prescription Drug Monitoring Programs (PDMPs) are operating in</div><div style="left: 90px; top: 425.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.979346);" data-canvas-width="393.10941666666656">49 states and several U.S. territories. Current methods for surveillance</div><div style="left: 90px; top: 442.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.995478);" data-canvas-width="393.36299999999994">of prescription drug related behaviors, include the mean daily dosage</div><div style="left: 90px; top: 459.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.05006);" data-canvas-width="397.64133333333325">of morphine milligram equivalent (MME) per patient, annual</div><div style="left: 90px; top: 475.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.964983);" data-canvas-width="395.0800000000001">percentage of days with overlapping prescriptions per patient, and</div><div style="left: 90px; top: 492.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02181);" data-canvas-width="394.4679999999999">annual multiple provider episodes for multiple controlled substance</div><div style="left: 90px; top: 509.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.973022);" data-canvas-width="348.58500000000015">prescription drugs per patient that are described elsewhere.</div><div style="left: 438.762px; top: 509.417px; font-size: 8.5px; font-family: serif; transform: scaleX(1.0176);" data-canvas-width="10.803499999999998">1,2</div><div style="left: 449.573px; top: 509.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.898314);" data-canvas-width="35.34866666666666">This</div><div style="left: 90px; top: 525.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00625);" data-canvas-width="396.311083333333">work builds on these efforts by extending longitudinal methods</div><div style="left: 90px; top: 542.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.960186);" data-canvas-width="395.0686666666669">to prescription drug behavior surveillance in order to predict risks</div><div style="left: 90px; top: 559.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00085);" data-canvas-width="213.64750000000004">associated with prescription drug use.</div><div style="left: 90px; top: 590.758px; font-size: 14.1667px; font-family: sans-serif; transform: scaleX(1.07287);" data-canvas-width="58.23916666666666">Methods</div><div style="left: 105px; top: 605.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.999366);" data-canvas-width="378.4766666666666">Schedule II prescription opioids from January 1, 2014 to February</div><div style="left: 90px; top: 622.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.982463);" data-canvas-width="395.58858333333325">29, 2016 from the Kansas Tracking and Reporting of Controlled</div><div style="left: 90px; top: 639.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.972053);" data-canvas-width="392.87141666666645">Substances (KTRACS) was used for this analysis. Prescription opioids</div><div style="left: 90px; top: 655.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.986777);" data-canvas-width="393.0824999999999">were linked to the 2016 version of the morphine milligram equivalent</div><div style="left: 90px; top: 672.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.996303);" data-canvas-width="395.73166666666646">conversion table from the National Center for Injury Prevention</div><div style="left: 90px; top: 689.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.0231);" data-canvas-width="72.36758333333333">and Control.</div><div style="left: 162.327px; top: 689.417px; font-size: 8.5px; font-family: serif;">3</div><div style="left: 166.608px; top: 689.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03537);" data-canvas-width="317.9779166666666">Population estimates were based on the 2015 County</div><div style="left: 90px; top: 705.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01592);" data-canvas-width="394.12516666666664">Vintage single-year of age bridged-race estimates from the National</div><div style="left: 90px; top: 722.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.978503);" data-canvas-width="393.1136666666663">Center for Health Statistics and used to calculate age-adjusted rates. A</div><div style="left: 90px; top: 739.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.982892);" data-canvas-width="393.0258333333335">daily high dose opioid prescription was defined as having greater than</div><div style="left: 90px; top: 755.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.989114);" data-canvas-width="393.2057500000001">or equal to 90 morphine milligram equivalent. Since this is a unit-day</div><div style="left: 90px; top: 772.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02662);" data-canvas-width="394.34333333333336">measure with patients experiencing multiple daily high dose opioid</div><div style="left: 90px; top: 789.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.99079);" data-canvas-width="395.67216666666656">days, the Prentice, William, and Peterson (PWP) recurrent event</div><div style="left: 90px; top: 805.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00748);" data-canvas-width="393.79224999999997">model was used to estimate the number of high-dose opioid days for</div><div style="left: 90px; top: 822.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01546);" data-canvas-width="244.6583333333334">Kansas patients by gender and age groups.</div><div style="left: 334.657px; top: 822.751px; font-size: 8.5px; font-family: serif; transform: scaleX(1.00078);" data-canvas-width="10.624999999999998">4,5</div><div style="left: 345.282px; top: 822.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03214);" data-canvas-width="138.84041666666667">Start time was the first</div><div style="left: 90px; top: 839.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00988);" data-canvas-width="393.86875">prescription date with a high-dose opioid and stop time was the next</div><div style="left: 90px; top: 855.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.953604);" data-canvas-width="394.96666666666687">high-dose opioid date during a study period from January 1, 2014</div><div style="left: 90px; top: 872.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.04567);" data-canvas-width="394.34049999999985">to Feb 29, 2016. The PWP model is a statistical model that allows</div><div style="left: 90px; top: 889.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.946868);" data-canvas-width="394.79666666666645">for the estimation of covariates on an event history (i.e. total time</div><div style="left: 90px; top: 905.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03167);" data-canvas-width="394.62949999999995">with prescription opioids, specifically high-dose opioids). Analysis</div><div style="left: 90px; top: 922.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00627);" data-canvas-width="396.0858333333337">was completed with a stratified Cox-proportional hazard model,</div><div style="left: 90px; top: 939.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02172);" data-canvas-width="396.5278333333328">sandwich covariance for dependent observations, and statistical</div><div style="left: 90px; top: 955.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.0394);" data-canvas-width="394.33200000000005">significance was assessed with a Wald Chi-square. PROC PHREG</div><div style="left: 90px; top: 972.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02821);" data-canvas-width="394.026">in SAS/STAT(R) 14.1 was used since it has a new FAST option for</div><div style="left: 90px; top: 989.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00408);" data-canvas-width="323.02833333333325">fitting large proportional counting process hazard model.</div><div style="left: 90px; top: 1020.76px; font-size: 14.1667px; font-family: sans-serif; transform: scaleX(1.08488);" data-canvas-width="51.17">Results</div><div style="left: 105px; top: 1035.71px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01276);" data-canvas-width="381.5253333333333">The age-adjusted rate of daily high-dose opioid patients was</div><div style="left: 90px; top: 1052.38px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.966496);" data-canvas-width="395.37750000000005">3.2 patients per 100 Kansas population-year (95% CI: 3.1 – 3.2).</div><div style="left: 90px; top: 1069.04px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03357);" data-canvas-width="394.4963333333333">Kansas patients aged 85 and older had the highest age-specific rate</div><div style="left: 90px; top: 1085.71px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.04259);" data-canvas-width="394.3249166666666">of 11.7 (95% CI: 11.5 –11.9). Preliminary recurrent event analysis</div><div style="left: 90px; top: 1102.38px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00373);" data-canvas-width="396.355">shows on average nearly a quarter of approximately 50 million</div><div style="left: 90px; top: 1119.04px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03399);" data-canvas-width="394.3305833333336">Schedule II opioid patient days were high-dose opioid patient days</div><div style="left: 90px; top: 1135.71px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02298);" data-canvas-width="394.40849999999995">among 785,514 Kansan patients with any prescribed opioid history.</div><div style="left: 90px; top: 1152.38px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02059);" data-canvas-width="394.13650000000007">In an initial result stratified by the number of high-dose opioid days</div><div style="left: 90px; top: 1169.04px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01435);" data-canvas-width="394.00616666666645">and adjusting only for age, males on average had approximately 7%</div><div style="left: 90px; top: 1185.71px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01128);" data-canvas-width="394.0274166666666">higher hazard of recurrent Schedule II high-dose opioid prescription</div><div style="left: 90px; top: 1202.38px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03651);" data-canvas-width="113.6875">days than females (</div><div style="left: 203.661px; top: 1205.07px; font-size: 14.1667px; font-family: sans-serif;">β</div><div style="left: 211.501px; top: 1202.38px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03645);" data-canvas-width="273.1049999999998">: 0.07, S.E: 0.002, p<0.0001). Kansas patients</div><div style="left: 510px; top: 312.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02203);" data-canvas-width="394.2158333333333">aged 45 to 54 compared to Kansas patients 85 and older on average</div><div style="left: 510px; top: 329.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00509);" data-canvas-width="390.02533333333326">had approximately 14% higher hazard of recurrent Schedule II high-</div><div style="left: 510px; top: 345.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00069);" data-canvas-width="174.70333333333335">dose opioid prescription days (</div><div style="left: 684.69px; top: 348.4px; font-size: 14.1667px; font-family: sans-serif;">β</div><div style="left: 692.465px; top: 345.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00104);" data-canvas-width="168.15833333333333">: 0.14, S.E: 0.007, p<0.0001).</div><div style="left: 510px; top: 377.425px; font-size: 14.1667px; font-family: sans-serif; transform: scaleX(1.10336);" data-canvas-width="85.01416666666667">Conclusions</div><div style="left: 525px; top: 392.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03533);" data-canvas-width="381.8639166666667">This work demonstrates the application of survival analysis</div><div style="left: 510px; top: 409.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.03559);" data-canvas-width="394.59549999999984">techniques to estimate the population at risk for high-dose opioids,</div><div style="left: 510px; top: 425.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.96189);" data-canvas-width="392.8841666666666">which varies by the length of the total opioid prescription history. Early</div><div style="left: 510px; top: 442.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.955625);" data-canvas-width="394.5615">results from the recurrent event analysis showed that Kansas male</div><div style="left: 510px; top: 459.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.00425);" data-canvas-width="393.68316666666664">and patients aged 45 to 54 years had the longest history of high-dose</div><div style="left: 510px; top: 475.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.996558);" data-canvas-width="393.32475000000005">opioids. Annual cross-sectional population estimates may incorrectly</div><div style="left: 510px; top: 492.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.963741);" data-canvas-width="394.9411666666665">estimate the estimated risk of high-dose prescription opioids since</div><div style="left: 510px; top: 509.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.961585);" data-canvas-width="395.0672500000002">it assumes all patients have the same prescription history. PDMPs</div><div style="left: 510px; top: 525.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.02291);" data-canvas-width="394.4849999999995">are longitudinal databases. Survival analysis methods like recurrent</div><div style="left: 510px; top: 542.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.980109);" data-canvas-width="393.03999999999985">event models can leverage the longitudinal structure to more precisely</div><div style="left: 510px; top: 559.043px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.01478);" data-canvas-width="394.17333333333335">estimate risk statistics. Future work includes incorporation of health</div><div style="left: 510px; top: 575.709px; font-size: 14.1667px; font-family: serif; transform: scaleX(0.990087);" data-canvas-width="393.13066666666674">outcomes data and further prescription covariates to assess the timing</div><div style="left: 510px; top: 592.376px; font-size: 14.1667px; font-family: serif; transform: scaleX(1.0008);" data-canvas-width="240.02583333333342">and intensity of opioid potency escalation.</div>}, number={1}, journal={Online Journal of Public Health Informatics}, author={Xiong, Fan}, year={2017}, month={May} }