The trial, led by Joseph Broderick of the Department of Neurology, used a computer search tree algorithm called CART to analyze CT scans of patients with strokes and the drug tPA to potentially treat strokes. It stands for Tissue Plasminogen Activator and breaks down blood clots. The key measurement was the NIHSS scale, which is a scale used to measure stroke-related neurological deficit. The scale is commonly used because it is most often accurate and predicts the effect of the stroke on the patient. The experiment tested patients on that scale after a certain time period, depending on the group being tested.
The research was done in several different trials. There were examinations and CT scans performed on patients to see the effectiveness of the tPA after 2 hours, 24 hours, 7-10 days, and 3 months. The first analysis consisted of the 2 hour and 24-hour examinations and were conducted in order to find the early effects of tPA activity. The second analysis consisted of the 7-10 day and 3-month trials and targeted its attention more towards long term outcomes of tPA usage. The key result was the trial that had the best outcome, which was after 24 hours and had an NIHSS (National Institute of Health Strokes Scale) of ≤ 2. The sample size of this trial was fifty-eight per group.
The trials were advantageous because they helped effectively determine when tPA was effective in treating stroke patients. Additionally, they used CT scans, which are an effective method of analysis. They are effective because they allow doctors to see where the bleeding in the brain is, so that they can figure out how to repair it. This is advantageous because it uses an existing treatment to try and gauge the effectiveness of a new treatment. Another advantage is that the methods used here, although having some problems and difficulties, played well with binary cut points. Binary cut points work best when the method is well-skewed and not evenly distributed, meaning the best results are those with clusters at either extreme. Treatment methods like the Barthel Index, Rankin Scale, and NIHSS are all very well skewed and data from these methods all show clusters at the extremes (most positive and most negative) and hardly any in the median (the middle).
Disadvantages of the research conducted here lie within the method that the researchers chose to use and the scans. The CART method is only capable of identifying possible binary points. This means that any possible nonbinary points would be excluded entirely. Another minor disadvantage was that the researchers used CT scans rather than MRI scans. CT scans are overall organ imaging, whereas MRI scans are more detailed and give the electrical activity in the scans. Even though MRI scans would have given more detailed information to them, researchers opted for CT scans, because they are much more time efficient and give results more quickly. MRI scans take longer but would have given researchers detailed effects of the drugs.
APA Citation:
Broderick, J. P., Joseph P. Broderick From the Department of Neurology, Lu, M., Mei Lu From the Department of Neurology, Kothari, R., Rashmi Kothari From the Department of Neurology, Levine, S. R., Steven R. Levine From the Department of Neurology, Lyden, P. D., Patrick D. Lyden From the Department of Neurology, Haley, E. C., E. Clark Haley From the Department of Neurology, Brott, T. G., Thomas G. Brott From the Department of Neurology, Grotta, J., James Grotta From the Department of Neurology, Tilley, B. C., Barbara C. Tilley From the Department of Neurology, Marler, J. R., … Broderick, C. to J. (2000, October 1). Finding the most powerful measures of the effectiveness of tissue plasminogen activator in the NINDS TPA stroke trial. Stroke. Retrieved October 12, 2021, from https://www.ahajournals.org/doi/full/10.1161/01.STR.31.10.2335
