Prognostic significance of risk stratifiers of mortality, including T wave alternans, after acute myocardial infarction: results of a prospective follow-up study.

INTRODUCTION: Occurrence of sustained microvolt-level T wave alternans (TWA) at a specified heart rate has been suggested to predict life-threatening arrhythmic events, but its prognostic value has not been well established in patients who survived an acute myocardial infarction (AMI). The purpose of this prospective study was to assess the predictive significance of various noninvasive risk indicators of mortality, including TWA, in consecutive post-AMI patients with optimized medical therapy. METHODS AND RESULTS: In addition to a symptom-limited predischarge exercise test with measurement of TWA, mortality risk was assessed using heart rate variability, 24-hour ECG recordings, baroreflex sensitivity, signal-averaged ECG, QTc interval, QT dispersion, and echocardiographic wall-motion index in 379 consecutive patients. Twenty-six patients (6.9%) died during a mean follow-up of 14 +/- 8 months. Sustained TWA was found in 56 patients (14.7%), none of whom died. Several risk variables, e.g., incomplete TWA test (inability to perform the exercise test or reach the required target heart rate of 105 beats/min), increased QRS duration on signal-averaged ECG, increased QT dispersion, long QTc interval, nondiagnostic baroreflex sensitivity result, and low wall-motion index, predicted all-cause mortality in univariate analyses. In multivariate analysis, the incomplete TWA test was the most significant predictor of cardiac death (relative risk 11.1, 95% confidence interval 2.4 to 50.8; P < 0.01). CONCLUSION: Sustained TWA during the predischarge exercise test after AMI does not indicate increased risk for mortality. An incomplete TWA test and several common risk variables provided prognostic information in this post-AMI population.

Cellular potassium depletion predisposes to hypokalaemia after oral sodium phosphate.

BACKGROUND: Oral sodium phosphate has become an attractive alternative to polyethylene glycol for colonic cleansing preparatory to elective colorectal surgery. Its use, however, has been associated with hypokalaemia. The authors of the present study tested the hypothesis that patients with cellular depletion of potassium are at significant risk for hypokalaemia with oral sodium phosphate bowel preparation. METHODS: In 23 patients, total body potassium was measured by whole-body counting and intracellular water volume was measured by bioimpedance analysis before oral sodium phosphate bowel preparation. Patients were divided into those whose serum potassium fell to 3.5 mmol/L or lower (Group 1) and those whose did not after sodium phosphate treatment (Group 2). RESULTS: The fall in serum potassium concentration over the period of oral sodium phosphate administration was significantly negatively correlated with intracellular potassium concentration measured prior to administration (r = -0.65, P = 0.0009). In Group 1, serum potassium concentration fell from 4.1+/-0.1 (standard error of the mean (SEM)) mmol/L to 3.2+/-0.1 mmol/L (P < 0.0001) while in Group 2 there was no significant change in this concentration (4.0+/-0.1 vs 3.9+/-0.1 mmol/L) as a result of sodium phosphate treatment. Intracellular potassium concentration prior to administration of sodium phosphate was significantly lower in Group 1 (117+/-9 mmol/L vs 143+/-7 mmol/L, P < 0.05). CONCLUSIONS: Caution should be exercised when treating patients with oral sodium phosphate who are considered to be cellularly depleted of potassium. These patients are at risk of hypokalaemia after this treatment.

When Technical Rationality Fails: Thinking about Terminally Ill Patients.

The theory that thinking is modelled on the social activity of argument is investigated through the views of general practitioners about terminally ill patients. The social activity of general practice centres on the consultation, which the doctor manages by 'technical rationality'. But this is difficult when the patient is terminally ill. In that case technical rationality is seen to fail and rhetorical skills are invoked. GPs' thinking about such consultations can be described using an agonistic model based on a hierarchy of objectives, strategies and tactics. The objective of keeping patients comfortable and dignified is aimed at through three strategies, and a variety of rhetorical tactics is drawn on in thinking about these strategies.

General practitioners' strategies and tactics of communication with the terminally ill.

Interviews with 22 randomly selected general practitioners (GPs) investigated their communication with terminally ill patients. In interview analysis a conceptual distinction was drawn between objectives, strategies and tactics. When treating terminal patients, GPs expressed the objectives of keeping the patient comfortable, painfree, happy and maintaining dignity. A strategy is a plan and mode of approaching patients existing over an extended time period. Three strategies were described by GPs for use when interacting with terminally ill patients. These are characterized as 'try to disclose', 'let the patient decide' and 'avoid disclosing'. Tactics refer to behaviours used within a single consultation, as part of a strategy. Six tactics are described: evasion, denial, uncertainty, hints and prompts, euphemism and reassurance. Different strategies imply quite different forms of consultation. Thus to understand a consultation we must place it into the context of the series.

Homicide offenders in Auckland, New Zealand.

In the 14-year period from 1976 to 1989, 183 homicide offenders have been recorded in Auckland, New Zealand. Data accessed from police files show that Maori and Polynesians made up the majority of the offenders. Almost all offenders were male, and the largest proportion was between 20 and 24 years of age. Racial differences were noted in the methods used to commit the homicide, in the offender-victim relationship, and in whether the offender acted alone. The majority of offenders were not convicted of murder, but were convicted on lesser charges.

Child passenger fatalities and restraint use in Auckland.

AIMS: To determine the proportion of children killed as passengers in car crashes who were unrestrained and to estimate the reduction in mortality that might result from increasing child restraint utilisation. METHODS: All children killed as passengers in car crashes, over the period 1980-90, were identified from the Auckland coroners records. Data were recorded on sex, age, use and type of restraint, position within the motor vehicle, pattern of injuries and place of death. RESULTS: There were 49 children (25 male, 24 female) killed as passengers in car crashes over the study period (2.14/100,000/year). The median age was eight years (range 0-14 years). Of the 38 children for whom restraint use could be determined, 31 (81.6%) were unrestrained and seven (18.4%) were restrained. None of the children aged 0-2 years were restrained. Thirty-five children (71.4%) died at the scene of the crash, three (6.2%) died during transport to hospital, and 11 (22.4%) died in hospital. CONCLUSIONS: Based on overseas reports of the efficacy of child restraints, close to half (49%) of all child passenger deaths in Auckland could have been prevented with appropriate child restraint use.

Death certification in Auckland.

AIMS: to investigate the reasons behind general practitioner referral of cases of natural death to the coroner, and also the understanding of general practitioners of the relevant legislation. METHODS: as cases of natural death were referred to the coroner over a seven month period, the deceased's general practitioner was contacted by telephone and questioned regarding his/her knowledge of the death and willingness to sign a cause of death certificate. Based upon this interview, cases were divided according to whether their general practitioner could have signed the death certificate or not. RESULTS: there were 314 cases where there was a contactable general practitioner. In 257 instances, the general practitioner would not have signed the death certificate: 183 found the death unexpected, and 74 had not seen the patient recently enough. In 57 instances the general practitioner would have signed the death certificate, but did not because: 23 were unaware of the death, 13 could not be contacted, 14 did sign a death certificate but this was rejected by the medical referee, and seven would have signed but did not for miscellaneous reasons. CONCLUSION: a disturbing finding was the general lack of understanding of certification legislation in particular the concept that the patient must have been seen in their last illness to be able to sign the death certificate. When signing death certificates, doctors should be thinking in terms of last illness rather than a vague period of time between two weeks and three months.

Differences between terminally ill patients who know, and those who do not know, that they are dying.

Four GPs each selected 6 patients, 2 terminally ill with knowledge of diagnosis and prognosis, 2 terminally ill without knowledge, and 2 chronically ill. Semi-structured interviews were conducted with the patients, and the transcripts' content was analyzed. There were large group differences in the predicted direction in frequency of utterances that indicated knowledge, which both validated the content analysis and verified the GPs' capacity to categorize patients according to knowledge. Terminally ill patients with knowledge produced significantly longer transcripts than the other two groups. Groups did not differ significantly on state or trait anxiety scales. Differences due to demographic variables could not account for the main effects. There were significant associations that involved age, anxiety, and the dependent variables of the content analysis, especially among the terminally ill patients without knowledge.

Role ambiguity in general practice: the care of patients dying at home.

Semi-structured interviews were given to 22 GPs to investigate their perceptions and strategies when dealing with dying patients. There seemed no clear division in practice between policies of telling and not telling patients about their illness. Most patients find out eventually, but GPs differ in the extent to which they aim for open discussions of the matter. Over half perceived care of the dying as a demanding but satisfying fulfillment of their professional skills, but nearly a quarter found it difficult and with few compensating rewards. We conclude that while there is a well established ideal type corresponding to the curative role, there is another which is still in the process of development. This type generates a role for doctors which parallels the patient's dying role, as distinct from the sick role.

KIE: Twenty-two British general practitioners were given semi-structured interviews to investigate their perceptions and strategies when dealing with dying patients. On the question of telling patients about the nature of their illnesses, there appeared to be a continuum, rather than a dichotomy, of practice--most patients learn the information eventually, but physicians differ in the extent to which they aim for open discussion of the issue. Over half of the physicians perceived terminal care to be a demanding but satisfying fulfillment of their professional skills. Nearly a quarter, however, found it both difficult and unrewarding. The authors suggest that such physicians are guided by only a curative role model and have not yet succeeded in internalizing a developing role model for dealing with dying patients.

Communication between general practitioners and patients dying at home.

Semi-structured interviews were carried out with four general practitioners and some of their terminally ill patients in order to investigate how doctors solve the problem of communication with these patients about the outcome of the illness. Three of the doctors preferred not to give explicit information, or to talk about the outcome, even when they knew that the patient realised that he or she was dying. Within this constraint they developed different ways of coping with the problem of how to talk to the patient. We interpret this behaviour as an attempt to remain within the framework of rules and expectations provided by the traditional roles of doctor and patient, a framework that would be threatened by the doctor's acknowledgement of helplessness. The fourth doctor did tell his patients, and treated the problem as one of counselling patients to help them cope with their predicament. His role of healer was thus extended to include terminally ill patients.

KIE: Two psychologists present excerpts from and discuss semi-structured interviews they conducted with four Durham, England, general practitioners and several of their terminally ill patients. The object of the study was to investigate the ways in which physicians perceive and express the problems of communicating with their dying patients, and how their patients perceive these communications. Three of the doctors prefer not to reveal diagnosis and prognosis explicitly in terminal illness even in cases where they are aware that the patients realize their condition. Each has developed a strategy of relationship that conceals physician helplessness and supports hope. The fourth doctor actively elicits questions and assumes a counseling role in helping his patients deal with impending death.

Changes in the probability of stuttering following a stutter: a test of some recent models.

To test the models of stuttering proposed by Still and Sherrard (1976), six stutterers (two adults and four children between 12 and 14 years old) read 33 or more passages of prose of approximately 200 words, in sessions of 16 or 17 passages. Predictions tested concerned changes in stuttering probability as words elapse following the last stutter. Such changes may be confounded with sequential dependencies among word classes, together with correlations between word classes and stuttering. To allow for this, a multiple regression was carried out on six independent variables--initial phoneme, grammatical class, length of word, position in sentence, session, and familiarity of passage. From this analysis, predicted changes in stuttering probability were generated, and compared with observed probabilities. There was a significant decrease in the difference between observed and expected probabilities over the first 30 words following a stutter. Also, two subjects (the adults) showed an initial increase in rate of stuttering. These results support a combination of the anxiety and feedback models of Still and Sherrard.

An analysis of individual differences in words stuttered.

Six stutterers (two adults and four children between 12 and 14 years old) read 33 or more passages of prose of approximately 200 words, in sessions of 16 or 17 passages. Words were classified by initial phoneme, grammatical class, length, and position in sentence, and proportions of stutters were examined as a function of these variables. The extensive data allowed a detailed analysis of individual differences, with the following results. Five subjects showed the usual higher rate of stuttering on consonants, though there were differences in which consonants were stuttered most, and one subject showed more stuttering on vowels. A markedly higher rate was found on initial words of sentences in two subjects, and these two subjects were also exceptional in stuttering more on content than function words, and in showing a type of stutter characterized by blocks rather than repetitions. Possible explanations for this pattern are discussed. Contrary to Taylor (1966) there were significant correlations between stuttering and grammatical class even when initial phoneme and word in sentence were held constant. There were no significant changes within sessions. Two subjects showed a significant decrease between sessions, one showed an increase; these changes were due to familiarity with the experimental situation, rather than with passages per se.

Location by odour and turn selection as two stages in the spontaneous alternation of rats.

Rats were given two free unrewarded trials per day in different T-mazes, oriented at 90 degrees to each other, and wiped with either Cajuput or Anise, which produced distinctive odours. The rate of spontaneous alternation was 82.8 per cent when the same odour was present on both trials, 62.5 per cent when the odour was changed between trials. This result was taken to support the hypothesis that alternation is a two-stage process, involving location then turn selection, and that odour is one of the stimuli used by the rat in locating itself.