Validation of the Hamilton Program for Schizophrenia Voices Questionnaire: Associations with emotional distress and wellbeing, and invariance across diagnosis and sex.

BACKGROUND: Voice-hearing is a transdiagnostic experience with evident negative effects on patients. Good quality measurement is needed to further elucidate the nature, impact and treatment of voice-hearing experiences across patient groups. The Hamilton Program for Schizophrenia Voices Questionnaire (HPSVQ) is a brief self-report measure which requires further psychometric evaluation. METHODS: Using data from a transdiagnostic sample of 401 adult UK patients, the fit of a conceptual HPSVQ measurement model, proposing a separation between physical and emotional voice-hearing characteristics, was tested. A structural model was examined to test associations between voice-hearing, general emotional distress (depression, anxiety, stress) and wellbeing. The invariance of model parameters was examined across diagnosis and sex. RESULTS: The final measurement model comprised two factors named 'voice severity' and 'voice-related distress'. The former comprised mainly physical voice characteristics and the latter mainly distress and other negative impacts. Structural model results supported voice-related distress as mediating the associations between voice severity and emotional distress and wellbeing. Model parameters were invariant across psychosis versus non-psychosis diagnosis and partially invariant across sex. Females experienced more severe and distressing voices and a more direct association between voice severity and general anxiety was evident. CONCLUSIONS: The HPSVQ is a useful self-report measure of voice-hearing with some scope for further exploration and refinement. Voice-related distress appears a key mechanism by which voice severity predicts general distress and wellbeing. Whilst our data broadly support interventions targeting voice-related distress for all patients, females may benefit especially from interventions targeting voice severity and strategies for responding.

When "privacy" threatens public health.

A European directive on privacy may make illegal much life-saving research on cancer in Britain, and the "fundamental principle" (no use of named records without permission of the individuals) could limit statistical work in Canadian cancer registries. Although the intent is to protect individuals from improper administrative use, this hardly applies to statistical aggregates that cannot identify anybody. Abstract position statements have done little to clarify the reasoning. To counter this, the present paper points to actual statistical uses of Canada's mortality database for individual follow-up on an unprecedented scale under conditions of strict confidentiality. The statistics serve to connect earlier causes with later harm. However, confusion still surrounds the question of whether privacy can be fundamentally violated by statistics drawn from personal records. Without wider understanding, real harm to health will compete poorly with imagined harm to privacy.

Age-related bias in probabilistic death searches due to neglect of the "prior likelihoods".

Major systematic bias can result when age-specific mortality data are sought through automated searches of large national death files, based on calculated odds in favor of a correct match. The bias occurs whenever the calculation fails to take into account: (a) the size of the death file being searched and (b) the age dependent likelihood that individual search record will find a correct match somewhere in that file. Indeed, there is widespread neglect of the "prior likelihood" or "prior odds" in favor of a fortuitous correct match on a single random pairing (i.e., before the identifiers have been compared). As a result, ratios of false matches to false nonmatches may be expected to vary widely in the different strata of a breakdown by age. A solution is described, using data from life tables and death statistics. The flexibility and merit of the "prior odds" for general application is emphasized.

Cohorts and privacy.

The extent to which major cohort studies now rely on centralized machine-readable files of personal records is taken largely for granted by investigators, but not by the public, privacy advocates, politicians, or the popular press. When widespread linkage into personal histories for statistical studies was first envisaged a few decades ago, it was regarded as impractical. But privacy advocates now fear the statistical uses even more than the administrative applications, perhaps because the latter are inevitable to prevent abuses, and often are associated with monetary benefits or personal convenience. Indeed a concept of 'privacy' has arisen that is deemed to be violated even in statistical studies where confidentiality is assured, unless the particular purpose is approved by all individuals. This concept motivates much of the political thinking. What could be lost to preventive medicine and clinical testing by this is illustrated with data from cohort studies reported in this journal (CCC) in 1993.

Distinguishing individual linkages of personal records from family linkages.

Errors occur in the automated linkage of medical and other records into personal histories, when household and family identifiers get wrongly treated in the same way as the strictly personal identifiers. This is because overwhelming evidence that two records refer to the same household or family, does not mean they necessarily represent the same person. Current practice in probabilistic record linkage fails to make this distinction, and combines indiscriminately the calculated odds from the mix of identifiers pertaining to all three sorts of entity. A tactic is described that avoids the resulting problem of mistaken personal identity. It does so by first converting composite odds from comparisons of household and family particulars, to a magnitude that contributes correctly to the overall evidence in favour of a strictly personal linkage.

Application of exact ODDS for partial agreements of names in record linkage.

Automated methods for linking records pertaining to the same individuals have in the past made only crude use of the name information. A perceptive filing clerk is more sophisticated because humans retain a lifetime memory of instances in which variant forms of names were employed interchangeably, and of synonyms that sometimes did not even resemble one another. This limitation of the machine can be rectified, but the body of knowledge required to serve as its memory must be large. The needed data have now been brought together on a suitable scale, from many past searches of Canada's Mortality Data Base. Described here are the development and use of the resulting tables of essentially exact discriminating powers (or ODDS) to do with comparisons of male given names. The aim is to reduce the proportion of ambiguously linked pairs of records requiring labor intensive clerical resolution. The tables are intended for general use in this country, and as a model for similar facilities appropriate to other populations.

Discriminating powers of partial agreements of names for linking personal records. Part I: The logical basis.

Machines have difficulty when using people's names to link medical and other records pertaining to the same individuals because of nicknames, ethnic synonyms, truncations, misspellings and typographical errors. Present algorithms used to compute the discriminating powers (or ODDS) associated with partial agreements of names are based, inappropriately, on the degrees of outward similarity alone. They are particularly ineffective in dealing with names that look alike but are unrelated, and with related names that have little apparent similarity. A fundamentally different rationale is, therefore, proposed which, like the human mind, assesses the relatedness of two alternative forms of a name in terms of how often they are used, interchangeably in practice. This must be taken into account if the associated discriminating powers (ODDS) are to be correctly computed. A way of implementing this more precise approach is described and illustrated, using the given names on linked records from an earlier epidemiological study. This first study of two describes the logical basis for record linkage, a second one the empirical test.

Discriminating powers of partial agreements of names for linking personal records. Part II: The empirical test.

The preceeding paper examined the logical basis of an exact way of calculating the discriminating powers of people's names when they only partially agree. The method has application to automated file searching and record linkage. The present account describes an empirical test of the approach. Use is made of some 2000 comparison pairs of male given names, obtained as a byproduct from an earlier linkage study. The test shows that exact value-specific ODDS can indeed be calculated for common names when compared with their accepted synonyms (e.g. JOSEPH versus JOE). Moreover, the use can be extended to include rare variants, by arranging these into groups defined in value-specific terms (e.g. as selected blocks in an alphabetically sequenced listing, or combinations of such blocks). A majority of all name comparisons may be handled in this manner. The added precision serves to reduce the numbers of records that are ambiguously linked and require labour intensive clerical resolution.

Genetic disorders in children and young adults: a population study.

The data base of an ongoing population-based registry with multiple sources of ascertainment was used to estimate the present population load from genetic disease in more than 1 million consecutive live births. It was found that, before approximately age 25 years, greater than or equal to 53/1,000 live-born individuals can be expected to have diseases with an important genetic component. This total was composed of single-gene disorders (3.6/1,000), consisting of autosomal dominant (1.4/1,000), autosomal recessive (1.7/1,000), and X-linked recessive disorders (0.5/1,000). Chromosomal anomalies accounted for 1.8/1,000, multifactorial disorders (including those present at birth and those of onset before age 25 years) accounted for 46.4/1,000, and cases of genetic etiology in which the precise mechanism was not identified accounted for 1.2/1,000. Previous studies have usually considered all congenital anomalies (ICD 740-759) as part of the genetic load, but only those judged to fit into one of the above categories were included in the present study. Data for congenital anomalies are therefore also presented separately, to facilitate comparison with earlier studies. If all congenital anomalies are considered as part of the genetic load, then greater than or equal to 79/1,000 live-born individuals have been identified as having one or other genetic disorder before approximately age 25 years. These new data represent a better estimate of the genetic load in the population than do previous studies.

Lung cancer mortality (1950-80) in relation to radon daughter exposure in a cohort of workers at the Eldorado Port Radium uranium mine: possible modification of risk by exposure rate.

A cohort study of 2,103 workers employed between 1942 and 1960 at a uranium mine in the Northwest Territories, Canada, was conducted. A total of 57 lung cancer deaths was observed (expected = 24.73, P less than .0001). There was a highly significant linear relationship between exposure and increased risk of lung cancer, giving estimates for the relative and attributable risk coefficients of 0.27 per working level month (WLM) and 3.10 per WLM per 10(6) person-years. These risk coefficients were substantially less than those estimated from the experience of miners in the Beaverlodge mine, which have previously been reported. Any biases in the present estimates are likely to have been upward, and therefore they probably represent an upper limit. The major difference between the two mine cohorts is in the exposure rate, since the Port Radium miners were exposed to much greater concentrations of radon daughters than the Beaverlodge miners. It is postulated that risk of lung cancer from radon daughter exposure may be modified by exposure rate, for which hypothesis there is some support from other epidemiologic data.

Lung cancer mortality (1950-80) in relation to radon daughter exposure in a cohort of workers at the Eldorado Beaverlodge uranium mine.

A cohort study of 8,487 workers employed between 1948 and 1980 at a uranium mine in Saskatchewan, Canada, has been conducted. A total of 65 lung cancer deaths was observed (34.24 expected, P less than 10(-5)). There was a highly significant linear relationship between dose and increased risk of lung cancer giving estimates for the relative and attributable risk coefficients of 3.28% per working level month (WLM) and 20.8 per WLM per 10(6) person-years. Age at first exposure had a significant modifying effect on risk. The interaction of exposure with age at observation fits a relative risk model well. The similarity of these results to a recent study of Swedish iron miners with similar levels of relatively low exposure suggests that exposure to radon daughter products may be a major contributory factor to lung cancer occurring among nonsmokers in the general population. The results also reinforce concerns as to the appropriateness of present occupational exposure standards.

Reliability of computerized versus manual death searches in a study of the health of Eldorado uranium workers.

An epidemiological follow-up study of 16,000 uranium mine and refinery employees has made use of computerized techniques for searching a national death file. The accuracy of this computerized matching has been compared with that of corresponding manual searches based on one-eighth of the worker file. The national death file--Canadian Mortality Data Base--at Statistics Canada includes coded causes of death for all deaths back to 1950. The machine search was carried out using a generalized record linkage system based upon a probabilistic approach. The machine was more successful than the manual searchers and was also less likely to yield false linkages with death records not related to the study population. In both approaches accuracy was strongly dependent on the amount of personal identifying information available on the records being linked.

Automated follow-up facilities in Canada for monitoring delayed health effects.

Increased public awareness of the possible presence of carcinogens and other potentially harmful agents in the workplace and in other areas of the environment has created a demand for studies to determine the extent of the risks associated with exposure to such agents. These studies require large numbers of individuals in various "control" populations to be followed-up over long periods of time. Such large-scale follow-up can be greatly facilitated where information on all deaths and on cases of serious morbidity is accumulated centrally, for a whole country, in a form that permits rapid searching by computer and in which individuals are well enough identified to minimize the possibility of mistaken identity. The Canadian Mortality Data Base and the National Cancer Incidence Reporting System are two such centralized follow-up facilities which have been developed in Canada. We describe here the manner in which these files are used, the problems encountered, and their solutions.