A recent study that assessed the risk of overdose in GS1101 subjects who had continuously received opioids for chronic non-cancer pain for at least three months found overdoses among 0.5% of subjects (1.5 events/1,000 person years of follow-up), and observed that subjects receiving 100 mg or more of morphine had a nine-fold increase in overdose risk [17]. In the present study approximately Inhibitors,research,lifescience,medical twenty percent of the subjects received such doses, at some time during treatment. Patients receiving these doses need close supervision. Future research should attempt to better define the patient groups in which prescription of opioids for non-cancer pain

is safe, and those in which it is not. Among the limitations of the present study, are: the PharMetrics Patient-Centric database reflects the commercially insured population and, therefore, overrepresents the healthy population which is able to work and underrepresents Inhibitors,research,lifescience,medical individuals 65 years or older and disabled and low income populations which are insured through Medicare or Medicaid programs. These limitations could Inhibitors,research,lifescience,medical affect the generalizability of the study findings. Attrition of subjects because of loss of health insurance may also affect a study like the present one, that is based on claims data, but some reassurance about this possible source of bias is provided by the finding doses in the preceding six months were similar

for subjects who left the cohort and those who remained in it. In addition, the claims database reflects only those Inhibitors,research,lifescience,medical dispensings that were submitted for reimbursement in the outpatient setting.

It is not able to identify medications that a patient may have obtained outside the health care benefit system. However, the PharMetrics Patient-Centric database captures data from all retail and mail-order dispensing regardless of which health care provider in the plan issues the prescription. Pain intensity or pain relief information is absent in claims databases. Therefore, it is unknown whether subjects Inhibitors,research,lifescience,medical stopped treatment with opioids because of disease improvement, the lack of efficacy, or the lack of tolerability. Systematic reviews have found that up to one third of patients in clinical trials withdraw because of opioid adverse events and up to 18% of patients withdraw because of inadequate pain relief [3,5]. ADAMTS5 The inability to link with certainty the opioid dispensings directly to the diagnoses that they were intended to address, and the fact that PharMetrics uses the ICD-9 system to code the medical diagnosis, precluded the use of a pain taxonomy based on the pain mechanism. Thus, no attempts were made to compare opioid dose across pain categories, with the exception of the presence of cancer diagnosis. The pattern of dose escalation observed in this study could be a reflection of the success of the adequate patient selection and careful evaluation by health care providers when prescribing opioids.