Three programs for computing Dunn-Sidák critical values.

The Dunn-Sidák multiple comparison procedure is used to test hypotheses and construct confidence intervals for two or more a priori nonorthogonal contrasts for population means. The procedure uses a levels that are not available in conventional t tables. In 1977 Games prepared a table of two tailed critical values that simplifies using the procedure. Researchers are accustomed to doubling alpha in a two-tailed t table to obtain the critical value in one tail. Unfortunately, doubling alpha in Games's table always gives one tailed critical values that are too small. Computational routines using the t inverse function in Microsoft Excel 98, SAS, and SPSS are presented for obtaining Dunn-Sidák one- and two-tailed critical values for ant familywise error rate, number of contrasts, and error degrees of freedom.

Treatment of comminuted distal radial fractures with preliminary horizontal finger trap traction and a Roger-Anderson external fixation device.

Over a five-year period, 43 patients with comminuted distal radial fractures were treated with a Roger-Anderson external fixation device after the fracture was aligned in Strong's horizontal finger trap traction. Nineteen patients (21 wrists) were available for personal interview and radiographic follow-up. A 0.5-mm loss of radial height and an average loss of 2.4 degrees of palmar angulation presented. Range of motion (ROM) was excellent, stiffness was nonexistent or minimal in 81%, pain was nonexistent or minimal in 86%, and weakness of grasp was nonexistent or minimal in 81%. Complications were minimal; they included three pin tract infections, two of which required pin removal before they resolved. One patient fractured a pin that also required removal. Strong's horizontal finger trap traction and the Roger-Anderson external fixation device simplified the sometimes difficult treatment of this fracture. It seems to be most effective in young athletic individuals who have good bone stock and very comminuted fractures.

Geophysical variables and behavior: XXXVIII. Effects of ionized air on the performance of a vigilance task.

The effects of positively or negatively ionized air on the performance of a visual vigilance task were investigated. The ion level was 100,000 ions/cm3, +/- 10,000. Also investigated were the effects of task-induced stress and nonstress. 40 male college students were required to press a hand-held switch as soon as they detected a signal on a simulated cathode-ray display. Each subject monitored the display continuously for 3 hr. in a room specially equipped for ion exposure and for control of light and noise. All subjects received continuous identical false feedback about their performance. Half of the subjects (the stress group) were told that they would receive an electric shock when their performance fell below an acceptable level. The false feedback was programmed so subjects in this group believed that they were in danger of receiving a shock, but none were actually administered. Analysis indicated that the percentage of signals detected was higher in the presence of negatively ionized air than in the presence of positively ionized air. Also performance-contingent stress resulted in better performance than nonstress.

Changing patterns in tibial fractures resulting from skiing.

Of the 734 adult tibial fractures treated in Aspen, Colorado, from 1968-1978, follow-up results were obtained on 527 (72%). The fractures were analyzed with regard to rate of healing, age of patient, and location of fractures. There was a statistically significant increase in healing time with increasing age between ages 16-40. The location of tibial fracture (proximal, middle, distal) had no significant influence on rate of healing. The occurrence of open tibial fracture was low (3% of total) and in most instances wounds were benign (Class I). Compartment syndrome is rarely encountered in skiing fracture because of the relatively low energy involved. All tibial fractures occurring between 1982-1984 were reviewed and compared to the earlier series. This comparison confirmed the continual overall decline in rate of tibial fracture from skiing (from 12.9/100,000 skier days in 1968 to 2.9/100,000 skier days in 1983). This decline was primarily in spiral fractures (from 70% of total in 1968 to 50% in 1983). The percent of transverse fractures was unchanged (+/- 15%) while the percent of short oblique fractures increased from 11% to 33%. Properly adjusted and maintained safety equipment afforded reasonable protection from spiral fractures of the tibia, but a conceptual departure from the current release binding design would be required to protect against the benign moments that produced transverse and short oblique fracture. The current treatment philosophy is based on fracture type and the availability or experience with newer treatment modes.

Acute surgical repair of the skier's thumb.

Complete ulnar collateral ligament tears incurred during the period fall 1977 through spring 1979 were treated under regional anesthesia by a single group of surgeons in an identical fashion as outpatients. Of 123 thumbs repaired, 69 (59%) were available for follow-up examination. There were 34 women and 35 men with an average age of 34.5 years (range, 16-61 years). The follow-up period averaged 31.6 months (range, 16-46 months). Weakness of pinch was described as none or mild in 66 (96%) and significant in only three (4%). Stiffness was described as none or mild in 66 (96%), moderate in two (3%), and severe in one (1%). Pain was described as none or mild in 68 (99%) and moderate in one (1%). Sixty-two patients (90%) preferred outpatient surgery with regional anesthesia whereas only seven (10%) did not. Forty-five (65%) of the 69 patients who required surgery had used traditional poles with straps, suggesting the pole as the causative factor. However, only 20% of noninjured skiers during the 1981-1982 season were using traditional poles with straps, and the total number of complete ulnar collateral ligament repairs did not decrease with the same number of skiers. Therefore, the change in pole design has not decreased the incidence of total ulnar collateral ligament tears. The results of acute surgical repair as an outpatient with regional anesthesia are excellent; 96% of the patients were pleased with the results. If the lesion is overlooked, the results of secondary construction are not nearly as good.

Primary knee ligament repair--revisited.

There were 4710 knee sprains resulting from skiing in the four Aspen ski areas between 1976 and 1979. Twenty percent of the patients (942) had complete tears. Of these, 302 elected to remain in Aspen for treatment. All were treated by primary ligament repair without augmentation. These cases were evaluated an average of 42 months after injury. Patients with isolated tears of the medial collateral ligament were found to be doing well; virtually all of them had returned to preinjury activity levels. Thirty-six percent of the isolated anterior cruciate repairs were rated failures, and 43% of the combination ACL-MCL injuries had failed because of anterior cruciate deficiency. Twenty-nine percent of the ACL and ACL-MCL injuries had meniscal tears. Cases that included meniscectomy had a failure rate twice as great as those in which the meniscus was preserved. The results following repair of anterior cruciate tears were not acceptable, and augmentation was indicated. Primary repair of medial collateral ligament tears produced excellent results. Meniscal tears were frequent in association with ligament disruption. Ligament repairs were less satisfactory when meniscectomy was performed at the time of the repair.

Maintenance of vigilance by non-task-related stimulation in the monitoring environment.

The effects of non-task-related stimulation on the performance of a simple visual vigilance task were investigated. 40 subjects monitored a stimulated CRT display for two 90-min sessions. The independent variables were variable vs constant intensity lighting and variable vs constant intensity noise. Probability of detection was not significantly different for the noise conditions; however, probability of detection was significantly higher under variable than under constant intensity lighting.