Keith, Robert L.

Patients in whom fistula speech has developed after laryngectomy usually have to occlude their stoma with a finger to talk. This prevents the air that is needed to activate the fistula from being exhaled through the stoma. In order to free the hand, Blom and Singer devised a diaphram‐valved stomal device known as the tracheostoma valve. This device snaps into a flexible retention housing that must be carefully cemented to the skin. Some patients who have undergone near‐total laryngectomy (and others with tracheo‐esophageal voice prostheses) are unable to use their devices because the moisture and pressure around their stomas break the seal. Since air that is used to produce voice with the valve tends to lift the housing away from the stoma, a better housing design would assure that this motion would improve the seal.

Laryngo-esophagectomy was performed with a colon transplant to replace the excised esophagus. The patient did not develop functional “colonic” speech but does function adequately using an artificial larynx. Cineradiologic evidence reveals that the transplanted segment of colon permits development of a pseudoglottis to produce sound. He was capable of injecting air and forming the pseudoglottis, but he interfered with the production of colonic voice by retropositioning his tongue to close the hypopharynx.

This study investigated the practicality of developing short forms of the Porch Index of Communicative Ability (PICA). Stepwise regression procedures were performed using data from 222 complete administrations of the PICA to determine the accuracy of prediction of overall scores on the basis of data from either fewer subtests or fewer items. Results indicated that the PICA is highly redundant and that it may be possible to develop useful shorter forms of it.

Two short versions (SPICAs) of the Porch Index of Communicative Ability (PICA)were developed and investigated, with specific attention to their ability to predict standard PICA overall scores of patients with aphasia. Ninety-three patients were administered three versions of the test in randomized order. Results indicate that SPICAs predict overall scores by standard administration at acceptable confidence levels in approximately one-third of the time required for the administration of the standard PICA.