Andersohn, Schade, Suissa, & Garbe (2006) said that many researchers investigated the relationship between risk of myocardial infarction and Cyclooxygenase-2 (COX-2) inhibitors (targets an enzyme that is responsible for inflammation and pain), but not on the relationship between the risk of ischemic stroke and COX-2 inhibitors. Their objective for this study was to examine the connection between the risk of ischemic stroke and COX-2 inhibitors.In order to find the relationship between the risk of ischemic stroke and COX-2 inhibitor, Andersohn, Schade, Suissa, & Garbe (2006) used UK General Practice Research Database (GPRD) to gather patients who were at least 40 years old that were prescribed at least one nonsteroidal anti-inflammatory drugs (reduces pain, fever, and inflammation). Researchers narrowed down to 3094 patients who have an ischemic stroke and 11859 patients who have similar age, sex, and the year of diseases/problems. To determine the connection between specific drugs that have COX-2 inhibitors and risk of ischemic stroke, Andersohn, Schade, Suissa, & Garbe (2006) grouped nonsteroidal anti-inflammatory drugs (NSAIDs) into rofecoxib, celecoxib, etoricoxib, diclofenac, ibuprofen, naproxen, and other NSAIDs.The risk of ischemic stroke with each NSAIDs was predicted from odd ratios (compare the odds of exposure on cases and control groups) by a program called SAS PHREG program. Andersohn, Schade, Suissa, & Garbe (2006) used SAS PHREG program to find odd ratios (ORs) of patients who were prescribed with drugs that have COX-2 inhibitors for the first time, and the ratios of the risk for use of NSAIDs that contains COX-2. Since the patients were taking a prescription of NSAIDs, Andersohn, Schade, Suissa, & Garbe (2006) had to control patients who had the potential confounders that may affect ischemic stroke such as atrial fibrillation, hyperlipidemia, hypertension, coronary heart disease, heart failure, or cerebrovascular diseases. The result showed that from 469674 patients that were gathered on June 1, 2000, and October 31, 2004, 3156 patients experienced an ischemic stroke; however, Andersohn, Schade, Suissa, & Garbe (2006) had to exclude 62 people because no control was found. From 3094 case groups patients and 11859 control groups patients, 1060 patients from case groups and 3990 patients from control groups showed that they were having rofecoxib, etoricoxib, or celecoxib. Based on those patients, the outcome showed that only rofecoxib and etoricoxib increases the risk of ischemic stroke. The limitation of this study can be made that there were no potential confounders that are related to the lifestyle factors such as an amount of physical exercise, diet, or poor sleep. Andersohn, Schade, Suissa, & Garbe (2006) only looked at potential confounders that were related to diagnosis or diseases. Lifestyle factors can give big changes to patients who have potential confounders of ischemic stroke. For example, if the patient had the good amount of lifestyle factors such as physical activities, he/she most likely will have a low risk of cardiovascular diseases (atrial fibrillation, heart failure, or coronary heart diseases) that affects ischemic stroke (January et al., 2014). Another example can be found that the risk of hyperlipidemia can decrease when the patient maintains a healthy diet.