Considerations for the design, analysis and presentation of in vivo studies

OBJECTIVE

To describe, explain and give practical suggestions regarding important principles and key methodological challenges in the study design, statistical analysis, and reporting of results from in vivo studies.

CONCLUSIONS

Pre-specifying endpoints and analysis, recognizing the common underlying assumption of statistically independent observations, performing sample size calculations, and addressing multiplicity issues are important parts of an in vivo study. A clear reporting of results and informative graphical presentations of data are other important parts.

An examination of the effect of different methods of scoring pain after a total knee replacement on the number of patients who report unchanged or worse pain

We identified a group of patients from the Swedish Arthroplasty Register who reported no relief of pain or worse pain one year after a total knee replacement (TKR). A total of two different patient-reported pain scores were used during this process. We then evaluated how the instruments used to measure pain affected the number of patients who reported no relief of pain or worse pain, and the relative effect of potential risk factors. Between 2008 and 2010, 2883 TKRs were performed for osteoarthritis in two Swedish arthroplasty units. After applying exclusion criteria, 2123 primary TKRs (2123 patients) were included in the study. The Knee injury and Osteoarthritis Outcome Score (KOOS) and a Visual Analogue Scale (VAS) for knee pain were used to assess patients pre-operatively and one year post-operatively. Only 50 of the 220 patients (23%) who reported no pain relief on either the KOOS pain subscale or the VAS for knee pain did so with both of these instruments. Patients who reported no pain relief on either measure tended to have less pain pre-operatively but a higher degree of anxiety. Charnley category C was a predictor for not gaining pain relief as measured on a VAS for knee pain. The number of patients who are not relieved of pain after a TKR differs considerably depending on the instrument used to measure pain.

The Swedish Knee Arthroplasty Register: a review

We are entering a new era with governmental bodies taking an increasingly guiding role, gaining control of registries, demanding direct access with release of open public information for quality comparisons between hospitals. This review is written by physicians and scientists who have worked with the Swedish Knee Arthroplasty Register (SKAR) periodically since it began. It reviews the history of the register and describes the methods used and lessons learned. Cite this article: Bone Joint Res 2014;3:217-22.

Why the P-value culture is bad and confidence intervals a better alternative

In spite of frequent discussions of misuse and misunderstanding of probability values (P-values) they still appear in most scientific publications, and the disadvantages of erroneous and simplistic P-value interpretations grow with the number of scientific publications. Osteoarthritis and Cartilage prefer confidence intervals. This is a brief discussion of problems surrounding P-values and confidence intervals.

Can sulphonylurea addition to lifestyle changes help to delay diabetes development in subjects with impaired fasting glucose? The Nepi ANtidiabetes StudY (NANSY)

The Nepi ANtidiabetes StudY (NANSY) is a 5-year randomized, double-blind, placebo-controlled trial in Swedish primary care, examining whether the development of type 2 diabetes (T2D) and retinopathy (separately reported) would be delayed in 40- to 70-year-old subjects with impaired fasting glucose (IFG) who, in addition to lifestyle changes, were treated with either placebo or low-dosage sulphonylurea (SU) (1-mg glimepiride; Amaryl). Of 274 subjects (163 men, 111 women), 138 were allocated to placebo (46.0% men, 56.8% women) and 136 to glimepiride (54.0% men, 43.2% women). The primary endpoint was conversion to diabetes. Average follow-up time was 3.71 years; 96 subjects converted to diabetes, 55 allocated to placebo and 41 to glimepiride (absolute difference 9.8%; p = 0.072). In conclusion, the study failed to support the notion that low-dose SU added to lifestyle changes in IFG subjects would help to delay the conversion to diabetes.

Sampling uncertainty in medical research

Statistical analysis is ubiquitous in medical research, but fundamental statistical principles are not always well understood, which has negative effects for both authors and readers of scientific papers. The purpose of this brief review is to provide a methodological overview on populations, samples, analysis units and sampling uncertainty to facilitate an increased understanding of statistical concepts.

Femoral impaction bone grafting with the Exeter stem - the Swedish experience: survivorship analysis of 1305 revisions performed between 1989 and 2002

We identified 1305 femoral impaction bone grafting revisions using the Exeter stem performed between 1989 and 2002 in 30 hospitals throughout Sweden. There were 1188 patients with a mean age of 71 years (29 to 94) followed up for between five and 18 years. The participating departments reported 70 further revisions in total, of which 57 could also be identified on the Swedish National Arthroplasty Registry. Kaplan-Meier survivorship for all causes of failure was 94.0% (95% confidence interval (CI) 92 to 96) for women and 94.7% (95% CI, 92 to 96) for men at 15 years. Survivorship at 15 years for aseptic loosening was 99.1% (95% CI 98.4 to 99.5), for infection 98.6% (95% CI 97.6 to 99.2), for subsidence 99.0% (95% CI 98.2 to 99.4) and for fracture 98.7% (95% CI 97.9 to 99.2) Statistically significant predictors of failure were the year in which revision was conducted (p < 0.001). The number of previous revisions was slightly above the level of significance (p = 0.056). Age, gender, the length of the stem and previous septic loosening were not predictors of failure (p = 0.213, p = 0.399, p = 0.337, p = 0.687, respectively). The difference in survivorship between high- and low-volume departments was only 3% at ten years. We conclude that impaction bone grafting with the Exeter stem has an excellent long-term survivorship following revision arthroplasty. The technique of impaction grafting appears to be reliable, can be learned rapidly and produces a predictably low incidence of aseptic loosening.

Health-care quality registers: outcome-orientated ranking of hospitals is unreliable

Public disclosure of outcome-orientated ranking of hospitals is becoming increasingly popular and is routinely used by Swedish health-care authorities. Whereas uncertainty about an outcome is usually presented with 95% confidence intervals, ranking's based on the same outcome are typically presented without any concern for bias or statistical precision. In order to study the effect of incomplete registration of re-operation on hospital ranking we performed a simulation study using published data on the two-year risk of re-operation after total hip replacement. This showed that whereas minor registration incompleteness has little effect on the observed risk of revision, it can lead to major errors in the ranking of hospitals. We doubt whether a level of data entry sufficient to generate a correct ranking can be achieved, and recommend that when ranking hospitals, the uncertainties about data quality and random events should be clearly described as an integral part of the results.

Increased long-term mortality in patients less than 55 years old who have undergone knee replacement for osteoarthritis: results from the Swedish Knee Arthroplasty Register

Patients with osteoarthritis undergoing knee replacement have been reported to have an overall reduced mortality compared with that of the general population. This has been attributed to the selection of healthier patients for surgery. However, previous studies have had a maximum follow-up time of ten years. We have used information from the Swedish Knee Arthroplasty Register to study the mortality of a large national series of patients with total knee replacement for up to 28 years after surgery and compared their mortality with that of the normal population. In addition, for a subgroup of patients operated on between 1980 and 2002 we analysed their registered causes of death to determine if they differed from those expected. We found a reduced overall mortality during the first 12 post-operative years after which it increased and became significantly higher than that of the general population. Age-specific analysis indicated an inverse correlation between age and mortality, where the younger the patients were, the higher their mortality. The shift at 12 years was caused by a relative over-representation of younger patients with a longer follow-up. Analysis of specific causes of death showed a higher mortality for cardiovascular, gastrointestinal and urogenital diseases. The observation that early onset of osteoarthritis of the knee which has been treated by total knee replacement is linked to an increased mortality should be a reason for increased general awareness of health problems in these patients.

Prospective malignancy grading of invasive squamous carcinoma of the uterine cervix. Prognostic significance in a long-term follow-up

A multifactorial grading score (MGS) for invasive squamous cell carcinoma of the uterine cervix has demonstrated its capacity to predict survival in a 5-10 year perspective and metastasis frequencies, and is a valuable tool for treatment schedules. In this study it was shown that the power of prognosis is valid even up to 20 years. In this material from 619 cervical carcinoma patients the MGS scores turned out to remain as strong as earlier proven. Earlier studies have shown that MGS is superior to other mono- and multifactorial grading systems, histological differentiation into cell types, age, clinical stage, irradiation and DNA-analysis. Treatment of cervical squamous cell carcinoma is more specific today to meet the patients' need for instance to preserve fertility or to minimize operation and eventually radiotherapy. The MGS score is a strong prognostic tool in patients with cervical carcinoma.

Alcohol, tobacco and breast cancer--collaborative reanalysis of individual data from 53 epidemiological studies, including 58,515 women with breast cancer and 95,067 women without the disease

Alcohol and tobacco consumption are closely correlated and published results on their association with breast cancer have not always allowed adequately for confounding between these exposures. Over 80% of the relevant information worldwide on alcohol and tobacco consumption and breast cancer were collated, checked and analysed centrally. Analyses included 58,515 women with invasive breast cancer and 95,067 controls from 53 studies. Relative risks of breast cancer were estimated, after stratifying by study, age, parity and, where appropriate, women's age when their first child was born and consumption of alcohol and tobacco. The average consumption of alcohol reported by controls from developed countries was 6.0 g per day, i.e. about half a unit/drink of alcohol per day, and was greater in ever-smokers than never-smokers, (8.4 g per day and 5.0 g per day, respectively). Compared with women who reported drinking no alcohol, the relative risk of breast cancer was 1.32 (1.19-1.45, P<0.00001) for an intake of 35-44 g per day alcohol, and 1.46 (1.33-1.61, P<0.00001) for >/=45 g per day alcohol. The relative risk of breast cancer increased by 7.1% (95% CI 5.5-8.7%; P<0.00001) for each additional 10 g per day intake of alcohol, i.e. for each extra unit or drink of alcohol consumed on a daily basis. This increase was the same in ever-smokers and never-smokers (7.1% per 10 g per day, P<0.00001, in each group). By contrast, the relationship between smoking and breast cancer was substantially confounded by the effect of alcohol. When analyses were restricted to 22 255 women with breast cancer and 40 832 controls who reported drinking no alcohol, smoking was not associated with breast cancer (compared to never-smokers, relative risk for ever-smokers=1.03, 95% CI 0.98-1.07, and for current smokers=0.99, 0.92-1.05). The results for alcohol and for tobacco did not vary substantially across studies, study designs, or according to 15 personal characteristics of the women; nor were the findings materially confounded by any of these factors. If the observed relationship for alcohol is causal, these results suggest that about 4% of the breast cancers in developed countries are attributable to alcohol. In developing countries, where alcohol consumption among controls averaged only 0.4 g per day, alcohol would have a negligible effect on the incidence of breast cancer. In conclusion, smoking has little or no independent effect on the risk of developing breast cancer; the effect of alcohol on breast cancer needs to be interpreted in the context of its beneficial effects, in moderation, on cardiovascular disease and its harmful effects on cirrhosis and cancers of the mouth, larynx, oesophagus and liver.

Glycaemic control, disease duration and beta-cell function in patients with Type 2 diabetes in a Swedish community. Skaraborg Hypertension and Diabetes Project

AIMS

To examine determinants for glycaemic control in primary care patients with Type 2 diabetes.

METHODS

In a community-based surveillance of primary care patients with Type 2 diabetes, 190 men and 186 women were consecutively identified and examined for cardiovascular risk factors. Insulin resistance and beta-cell function were estimated using homeostasis model assessment (HOMA). Good glycaemic control was defined as HbA(1c) < 6.5%.

RESULTS

Following adjustment for age and gender, HbA(1c) > or = 6.5% was associated with duration of diabetes (10.6 vs. 6.4 years, P < 0.001), lower levels of serum insulin (6.3 vs. 8.0 mU/l, P = 0.012), higher serum triglyceride levels (2.0 vs. 1.7 mmol/l, P = 0.002) and impairment of beta-cell function (HOMA index 19.5 vs. 45.8, P < 0.001). The association between HbA(1c) levels and duration remained with adjustment for age, gender, waist-hip ratio (WHR) and serum triglycerides (odds ratio (OR) for HbA(1c) > or = 6.5% by 5 years diabetes duration = 1.7; 95% confidence interval (CI) 1.4--2.1) but was lost following additional adjustment for beta-cell function (OR for HbA(1c) > or = 6.5% = 1.3; 95% CI 0.96-1.7). In a separate linear regression with beta-cell function as the dependent variable there was a significant association with HbA1c after adjustments for differences in age, gender, WHR, serum triglyceride levels and diabetes duration (P < 0.001).

CONCLUSIONS

Increasing HbA1c by time was associated with declining beta-cell function.

The NEPI antidiabetes study (NANSY). 1: short-term dose-effect relations of glimepiride in subjects with impaired fasting glucose

AIM

NANSY is a randomised, placebo-controlled Swedish-Norwegian study which aims to include 2 x 1112 male and female subjects with impaired fasting glucose (IFG), to assess whether conversion to type 2 diabetes can be delayed by addition of sulphonylurea to dietary regulation and increased exercise. This pilot study was conducted to find the optimum dose of glimepiride in NANSY.

METHODS

In a double blind trial in primary care 25 IFG subjects were in random order exposed to single doses and one-week treatments with 0 (placebo), 0.5, 1.0 and 2.0 mg glimepiride once daily. The optimum dose was assessed by measuring blood glucose during oral 75 g glucose tolerance test (OGTT), comparing fasting blood glucose, and the area under the blood glucose curve (AUC), and by monitoring hypoglycaemic events.

RESULTS

With single doses, there was a clear dose-response relationship for the reduction in AUC, with a statistically significant difference only between placebo (mean 1981, 95% confidence intervals (CI) 1883-2078) and 2 mg glimepiride (mean 1763, 95% CI 1665-1861). However, following 1-week treatments, the only significant difference was between placebo (mean 1934, 95% CI 1856-2012) and 1 mg glimepiride (mean 1714, 95% CI 1637-1792). Correspondingly, the only statistically significant difference in fasting blood glucose day 7 was between placebo (5.87 mmol/l, 95% CI 5.68-6.05 mmol/l) and 1 mg glimepiride (5.42 mmol/l, 95% CI 5.21-5.62 mmol/l). Chemical hypoglycaemia was common but hypoglycaemic symptoms were rare and similar between the active doses, and easily countered by the subjects.

CONCLUSIONS

The sulphonylurea dose-effect curve may be bell-shaped, perhaps due to down regulation of sulphonylurea receptors during chronic exposure. Alternatively, the finding could be a rebound phenomenon, secondary to preceding hypoglycaemia. The optimum dose for NANSY was found to be 1 mg glimepiride.

Antidepressant medication and suicide in Sweden

OBJECTIVE

To explore a possible temporal association between changes in antidepressant sales and suicide rates in different age groups.

METHODS

A time series analysis using a two-slope model to compare suicide rates in Sweden before and after introduction of the selective serotonin reuptake inhibitors, SSRIs.

RESULTS

Antidepressant sales increased between 1977-1979 and 1995-1997 in men from 4.2 defined daily doses per 1000 inhabitants and day (DDD/t.i.d) to 21.8 and in women from 8.8 to 42.4. Antidepressant sales were twice as high in the elderly as in the 25-44-year-olds and eight times that in the 15-24-year-olds. During the same time period suicide rates decreased in men from 48.2 to 33.3 per 10(5) inhabitants/year and in women from 20.3 to 13.4. There was significant change in the slope in suicide rates after the introduction of the SSRI, for both men and women, which corresponds to approximately 348 fewer suicides during 1990-1997. Half of these 'saved lives' occurred among young adults.

CONCLUSION

We demonstrate a statistically significant change in slope in suicide rates in men and women that coincided with the introduction of the SSRI antidepressants in Sweden. This change preceded the exponential increase in antidepressant sales.

Attitudes toward xenotransplantation--patients waiting for transplantation versus the general public

The aim of this study was to survey attitudes toward xenotransplantation and to investigate whether there is a difference in attitudes between patients awaiting a kidney transplantation and the general public. A questionnaire was sent to randomly selected members of the public aged 18-75 (n = 1000) and to all patients of the same age group waiting for kidney transplants in Sweden in 1998 (n = 460). Among the public, 60 % expressed a positive attitude toward receiving an animal kidney graft with the same degree of risk as a human kidney graft, compared with 66 % for the patients. The proportion in favour of receiving a heart remained 60 % for the public, but rose to 70 % for the patients. If a human heart was not available, 61 % of the public were for the use of an animal heart, compared with 73 % in the patient group. A majority of the respondents would accept a transplant from an animal, provided the result and risk of infection were the same as with a human transplant. A greater proportion of patients had a positive attitude to receiving a xenotransplant than did the general public. A life threatening situation marginally increased the positive proportions.

Polymorphism in the beta(1)-adrenergic receptor gene and hypertension

BACKGROUND

The Arg389 variant of the beta(1)-adrenergic receptor gene mediates a higher isoproterenol-stimulated adenylate cyclase activity than the Gly389 variant in vitro. We investigated whether the Arg389Gly or the Ser49Gly polymorphism is associated with hypertension in Scandinavians. Methods and Results-- A total of 292 unrelated, nondiabetic, hypertensive patients and 265 unrelated healthy control subjects were included in a case-control association study. From 118 families, 102 nondiabetic sibling pairs without antihypertensive medication who were discordant for the Arg389Gly polymorphism were selected for a sibling study. Allele and genotype frequencies of the Arg389Gly and Ser49Gly polymorphisms were compared between hypertensive patients and normotensive control subjects. Blood pressure and heart rate were compared between carriers of the different genotypes. In the case-control study, the age- and body mass index-adjusted odds ratio for hypertension in subjects homozygous for the Arg389 allele was 1.9 (95% confidence interval, 1.3 to 2.7; P=0.0005) when compared with carriers of 1 or 2 copies of the Gly389 allele. The genotype-discordant sibling pair analysis revealed that siblings homozygous for the Arg389 allele had significantly higher diastolic blood pressures (79.4+/-9.9 versus 76.0+/-10.1 mm Hg; P=0.003) and higher heart rates (68.3+/-11.0 versus 65.1+/-9.4 bpm; P=0.02) than siblings carrying 1 or 2 copies of the Gly389 allele. The Ser49Gly polymorphism was not associated with hypertension.

CONCLUSION

Our data suggest that individuals homozygous for the Arg389 allele of the beta(1)-adrenergic receptor gene are at increased risk to develop hypertension.

β 2 -Adrenergic Receptor Gene Variation and Hypertension in Subjects With Type 2 Diabetes

Abstract —The aim of this study was to investigate whether polymorphisms in the β 2 -adrenergic receptor gene (5′LC-Arg19Cys, Arg16Gly, Gln27Glu) are associated with hypertension in patients with or without type 2 diabetes and with the blood pressure levels in normotensive sib pairs. The association study included 291 hypertensive patients without type 2 diabetes, 124 hypertensive patients with type 2 diabetes, and 265 healthy control subjects from Sweden. In addition, normotensive sib pairs that were discordant for the Arg16Gly (72 pairs) and Gln27Glu (40 pairs) polymorphisms were identified in type 2 diabetes families from Finland. Genotyping was performed using polymerase chain reaction–restriction fragment-length polymorphism analysis. Homozygous carriers of the Arg16 allele had a significantly increased odds ratio (OR) for hypertension in patients with type 2 diabetes (OR 2.14; 95% confidence interval [CI], 1.05 to 4.33), particularly among lean (body mass index<27 kg/m 2 ) patients (OR 3.47; 95% CI, 1.06 to 11.33). The Gln27 allele showed a weaker association to hypertension (OR 1.55; 95% CI, 1.00 to 2.41) and was found to be in linkage disequilibrium with the Cys19 allele of the 5′LC-Arg19Cys polymorphism. In the paired-sibling analysis, siblings with at least 1 copy of the Arg16 allele had higher systolic blood pressure ( P =0.049), and nondiabetic siblings had a higher body mass index ( P =0.026) than siblings homozygous for the Gly16 allele. These results indicate that the Arg16 allele of the β 2 -adrenergic receptor gene confers an increased risk for hypertension in subjects with type 2 diabetes and is associated with higher blood pressure levels and higher body mass index in sib pairs who are discordant for the polymorphism.

Impaired glucose metabolism and obesity in Swedish patients with borderline isolated systolic hypertension: Skaraborg Hypertension and Diabetes Project

AIM

To assess the prevalence of borderline isolated systolic hypertension (borderline ISH), and to examine its association with other cardiovascular risk factors.

METHODS

A cross-sectional community-based study was carried out in 1993-1994 in Skara, Sweden, including 1109 randomly chosen subjects > or = 40 years old. Normotension (NT) was defined as systolic blood pressure (SBP) < 140 and diastolic blood pressure (DBP) < 90 mmHg, borderline ISH as SBP 140-159 and DBP < 90 mmHg and hypertension (HT) as SBP > or = 160 or DBP > or = 90 mmHg or ongoing treatment.

RESULTS

The prevalence of borderline ISH (n = 203) by age was 4% in ages 40-49 years, 15% in ages 50-59 years, 28% in ages 60-69 years and 25% in ages 70-79 years. With borderline ISH as reference, normotensive subjects less often had fasting blood glucose > 5.5 mmol/l (odds ratio (OR): 0.4, 95% CI: 0.26-0.75), BMI > 27 kg/m2 (OR: 0.6, 95% confidence intervals (CI): 0.42-0.85) and known diabetes (OR: 0.4, 95% CI: 0.16-0.95). Hypertensive subjects more often had high density lipoprotein (HDL) cholesterol < 1.0 mmol/l (OR: 2.0, 95% CI: 1.35-2.99), a history of previous cardiovascular disease (CVD) (OR: 1.7, 95% CI: 1.01-2.72), known diabetes (OR: 2.4, 95% CI: 1.29-4.58) and microalbuminuria (men) (OR: 1.9, 95% CI: 1.15-3.11).

CONCLUSION

Borderline ISH is a common condition. It is associated with a more unfavourable risk factor profile than that of normotensive subjects concerning primarily glucose metabolism and obesity. The prevalence of known diabetes increased with the degree of hypertension.

Mortality in elderly men with low psychosocial coping resources using anxiolytic-hypnotic drugs

OBJECTIVE

Insufficient coping with stress may lead to increased susceptibility for disease and death. Use of anxiolytic-hypnotic drugs has been suggested as a coping strategy, and some opinions have proposed their use as preventive medication. The aim of this study was to estimate if use of anxiolytic-hypnotic drugs counters the increased mortality observed in individuals lacking other coping strategies such as emotional support and social participation.

METHODS

A population based cohort study with 10-year (1982/83-1993) survival analysis was performed in 491 men born in 1914, living in the Swedish city of Malmö.

RESULTS

Compared with men with a high level of psychosocial coping resources who did not use anxiolytic-hypnotic drugs, men with a low level of psychosocial coping resources had a higher risk of death irrespective whether they used anxiolytic-hypnotic drugs, RR = 1.7 (95% CI 1.1-2.6) or not RR = 1.8 (95%: 1.3-2.5).

CONCLUSION

Anxiolytic-hypnotic drugs do not seem to counter increased mortality in elderly men with low psychosocial coping resources.

Disaccordance between estimation of endometrial thickness as measured by transvaginal ultrasound compared with hysteroscopy and directed biopsy in breast cancer patients treated with tamoxifen

OBJECTIVE

To estimate endometrial thickness by transvaginal ultrasound technique (TVU) in patients treated with tamoxifen; and to compare the results with hysteroscopy and directed biopsy, and to relate the findings to curettage material.

METHODS

Fifty-four women, over 50 years of age at the time of their primary breast cancer operation and scheduled for adjuvant treatment of tamoxifen, were investigated. If the endometrium measured 8 mm or more (35 patients) hysteroscopy and directed biopsy was performed.

RESULTS

In the 35 patients, hysteroscopy and curretage revealed polyps in 18 cases (51.4%), proliferation in 1 case (2.9%) and atrophy in 16 cases (45.7%). In 54.3% (19 cases) of the patients the histopathological finding was a polyp or benign proliferation of the endometrium in agreement with the thickened endometrium seen with TVU. In the remaining 45.7% (16 cases) of the patients a thickened endometrium by transvaginal ultrasound was false positive as the hysteroscopic and subsequent histopathological diagnosis only showed atrofia of the uterine endometrium.

CONCLUSIONS

In almost half of the patients the TVU gave a misleading picture concerning the condition of the endometrium. Changes in the myometrium and related connective tissue could be misinterpreted by the ultrasound technique as changes in the endometrium. These results may confirm the theory that stimulation and changes by antiestrogenic treatment of the uterine tissue is not only restricted to the endometrium. In patients with thickened endometrium measured by TVU, a dilatation and curettage is not sufficient, as TVU has not the ability to differentiate between endometrial polyps and myometrial changes. A hysteroscopy and directed biopsy could therefore be performed.

Fraud in medical research: an international survey of biostatisticians. ISCB Subcommittee on Fraud

The characteristics of scientific fraud and its impact on medical research are in general not well known. However, the interest in the phenomenon has increased steadily during the last decade. Biostatisticians routinely work closely with physicians and scientists in many branches of medical research and have therefore unique insight into data. In addition, they have methodological competence to detect fraud and could be expected to have a professional interest in valid results. Biostatisticians therefore are likely to provide reliable information on the characteristics of fraud in medical research. The objective of this survey of biostatisticians, who were members of the International Society for Clinical Biostatistics, was to assess the characteristics of fraud in medical research. The survey was performed between April and July 1998. The participation rate was only 37%. We report the results because a majority (51%) of the participants knew about fraudulent projects, and many did not know whether the organization they work for has a formal system for handling suspected fraud or not. Different forms of fraud (e.g., fabrication and falsification of data, deceptive reporting of results, suppression of data, and deceptive design or analysis) had been observed in fairly similar numbers. We conclude that fraud is not a negligible phenomenon in medical research, and that increased awareness of the forms in which it is expressed seems appropriate. Further research, however, is needed to assess the prevalence of different types of fraud, as well as its impact on the validity of results published in the medical literature.

Factors predictive of bile duct stones in patients with acute calculous cholecystitis

The objective of the study was to investigate the accuracy of using preoperative data for the prediction of bile duct calculi in patients operated on for acute cholecystitis. 279 consecutive patients underwent cholecystectomy with peroperative cholangiography for acute calculous cholecystitis in the Department of Surgery, Lund University Hospital, between 1985 and 1991. The correlation between 13 preoperative clinical and laboratory variables, and the incidence of bile duct calculi was studied. Among the 13 variables tested, serum bilirubin concentration and serum gamma-glutamyltransferase activity were independent factors of predictive significance. These two factors were used for constructing an additive prognostic index for the presence of bile duct stones. Thus, three groups of patients could be identified having a 3, 7-9 or 59% risk of harboring bile duct stones, corresponding to a sensitivity of 61% and a specificity of 93% in predicting the presence of bile duct stones in the 'high-risk' group. Logistic regression analysis permits accurate preoperative identification of bile duct stones in patients with acute calculous cholecystitis.

Reduced mortality after severe head injury will increase the demands for rehabilitation services

PRIMARY OBJECTIVE

In 1989, a new therapy to reduce intracranial pressure in severely head-injured patients was introduced in Lund. The new treatment reduced mortality significantly. The present study describes the quality of life for the survivors.

METHODS AND PROCEDURES

The study includes 53 patients treated during 1989-1994, according to a new treatment protocol for increased intracranial pressure ('Lund concept' group). During 1982-1986, 38 patients were managed according to a protocol including high dose thiopentone ('Thiopentone' group). The two groups are compared regarding neurophysical and psychiatric symptoms as well as aspects regarding the patient's role, performance, interpersonal relationship, frictions, feelings and satisfaction in work, areas of social and leisure activities, and extended family.

RESULTS

Mortality was reduced from 47% to 8%, but the number of patients with a persistent vegetative state and/or remaining severe disability did not increase. However, the number of patients with persisting emotional and intellectual deficits increased significantly.

CONCLUSION

The new treatment regime has dramatically increased the number of survivors after severe head trauma. Although most patients have a favourable outcome, there are more patients with remaining sequelae and disabilities, and the demand for qualified rehabilitation has increased.

Analysis of Swedish male breast cancer family data: a simple way to incorporate a common sibling effect

Based on a population-based cohort study, Olsson et al. [1993] found significant evidence for elevated incidence of breast and ovarian cancers among female first-degree relatives of men with breast cancer. Using an extension of logistic regressive models we investigate whether, after allowing for multifactorial familial correlations, single locus segregation could be the cause of the elevated incidence in these families. The logit for a given sib in the class D logistic regressive model depends on the order in which affected sibs occur in a sibship. That makes the model less appropriate for the situation where a polygenic component or a common sibling environment may be present, as well as being computationally cumbersome. In this paper, we propose to use the proportion of siblings in a sibship who are affected to quantify a sibling correlation. That not only relaxes the interchangeability problem but also makes the model computationally efficient. We then use this modified class D logistic regressive model for our segregation analysis. Using the proportion of siblings in a sibship who are affected as a covariate resulted in a significantly higher likelihoods in all the models we investigated. We found evidence for a dominant Mendelian gene leading to early age of onset of breast and/or ovarian cancer. This could either be a germline mutation of BRCA2 or a mutation in a gene different from BRCA2.

Associations between smoking and beta-cell function in a non-hypertensive and non-diabetic population. Skaraborg Hypertension and Diabetes Project

AIMS

An increased risk for Type 2 diabetes in male and female smokers has been associated with insulin resistance. However, this might also be the result of an adverse effect on the beta-cell. The aim of the present study was to examine the association between smoking and beta-cell function.

METHODS

A community-based, cross-sectional observation study. In 1994, a randomized age-stratified sample of men and women aged > or = 40 years in the city of Skara, Sweden, were invited to a survey of cardiovascular risk factors. In all, 1,109 subjects participated (80%). After the exclusion of subjects with known hypertension or diabetes mellitus, 874 subjects remained to explore. Samples were drawn after an overnight fast. Lifestyle (smoking, physical activity, alcohol consumption) was assessed using a questionnaire. Insulin resistance and insulin secretion were estimated using the homeostasis model assessment (HOMA).

RESULTS

Cigarette smoking men (n = 101) had a lower HOMA beta-cell value (58.1), than never-smokers (n = 158, beta-cell value 90.1, P < 0.001). The difference remained with adjustments for age, body mass index, daily alcohol intake and physical exercise habits: 25.9 (95% confidence interval (CI) 9.7-38.8, P = 0.003). Correspondingly, in men the HOMA beta-cell value was lower in current smokers than in ex-smokers (difference 24.3, 95% CI 11.1-35.2, P < 0.001). In women, no significant difference appeared in beta-cell function vs. different smoking status. There was no association between smoking status and insulin resistance.

CONCLUSIONS

At least in men, smoking may interfere with beta-cell function. The prevention of Type 2 diabetes should include strategies to stop smoking.

[Prevalence for clinically proved carpal tunnel syndrome is 4 percent]

This article summarizes the results of a large-scale population-based study conducted to determine the prevalence of carpal tunnel syndrome in the Swedish general population. The study utilized a health questionnaires as well as clinical and electrophysiological examinations. Population prevalence rates of carpal tunnel syndrome, based on clinical diagnosis and electrophysiological criteria, were calculated. Obesity and specific work-related hand activities were shown to be risk factors for carpal tunnel syndrome.

The role of biostatistics in the prevention, detection and treatment of fraud in clinical trials

Recent cases of fraud in clinical trials have attracted considerable media attention, but relatively little reaction from the biostatistical community. In this paper we argue that biostatisticians should be involved in preventing fraud (as well as unintentional errors), detecting it, and quantifying its impact on the outcome of clinical trials. We use the term 'fraud' specifically to refer to data fabrication (making up data values) and falsification (changing data values). Reported cases of such fraud involve cheating on inclusion criteria so that ineligible patients can enter the trial, and fabricating data so that no requested data are missing. Such types of fraud are partially preventable through a simplification of the eligibility criteria and through a reduction in the amount of data requested. These two measures are feasible and desirable in a surprisingly large number of clinical trials, and neither of them in any way jeopardizes the validity of the trial results. With regards to detection of fraud, a brute force approach has traditionally been used, whereby the participating centres undergo extensive monitoring involving up to 100 per cent verification of their case records. The cost-effectiveness of this approach seems highly debatable, since one could implement quality control through random sampling schemes, as is done in fields other than clinical medicine. Moreover, there are statistical techniques available (but insufficiently used) to detect 'strange' patterns in the data including, but no limited to, techniques for studying outliers, inliers, overdispersion, underdispersion and correlations or lack thereof. These techniques all rest upon the premise that it is quite difficult to invent plausible data, particularly highly dimensional multivariate data. The multicentric nature of clinical trials also offers an opportunity to check the plausibility of the data submitted by one centre by comparing them with the data from all other centres. Finally, with fraud detected, it is essential to quantify its likely impact upon the outcome of the clinical trial. Many instances of fraud in clinical trials, although morally reprehensible, have a negligible impact on the trial's scientific conclusions.

Polymorphism in the angiotensin converting enzyme but not in the angiotensinogen gene is associated with hypertension and type 2 diabetes: the Skaraborg Hypertension and diabetes project

OBJECTIVE

To study the association between polymorphisms in the angiotensin converting enzyme (ACE) gene and angiotensinogen (AGT) gene and hypertension and/or type 2 diabetes in a community population.

PATIENTS AND METHODS

The insertion (I)/deletion (D) polymorphism of the ACE gene and the M235T polymorphism of the AGT gene were genotyped in 773 nondiabetic individuals with hypertension, 193 normotensive patients with type 2 diabetes, 243 patients with type 2 diabetes and hypertension, and in 820 normotensive control individuals identified in a community-based study.

RESULTS

The DD genotype was associated with hypertension in individuals less than 70 years [odds ratio (OR) = 1.54, confidence interval (CI) = 1.09-2.18] and remained so when patients with type 2 diabetes were excluded from the analysis (OR = 1.45, CI = 1.01-2.09). The strongest association was with the combination of type 2 diabetes and hypertension (OR = 2.19, CI = 1.09-4.38). There was no association with type 2 diabetes without hypertension. No association was observed between the M235T variant or the 3'-microsatellite polymorphism of the AGT gene and hypertension.

CONCLUSION

The D-allele of the ACE gene ID polymorphism increases susceptibility to hypertension, particularly when associated with type 2 diabetes. No association was observed between the M235T variant or 3'-microsatellite polymorphism of the AGT gene and hypertension.