Does the missing data imputation method affect the composition and performance of prognostic models?

We already showed the superiority of imputation of missing data [via Multivariable Imputation via Chained Equations [MICE] method] over exclusion of them; however, the methodology of MICE is complicated. Furthermore, easier imputation methods are available. The aim of this study was to compare them in terms of model composition and performance. Three hundreds and ten breast cancer patients were recruited. Four approaches were applied to impute missing data. First we adopted an ad hoc method in which missing data for each variable was replaced by the median of observed values. Then 3 likelihood-based approaches were used. In the regression imputation, a regression model compared the variable with missing data to the rest of the variables. The regression equation was used to fill the missing data. The Expectation Maximum [E-M] algorithm was implemented in which missing data and regression parameters were estimated iteratively until convergence of regression parameters. Finally, the MICE method was applied. Models developed were compared in terms of variables significantly contributed to the multifactorial analysis, sensitivity and specificity. All candidate variables significantly contributed to the MICE model. However, grade of disease lost its effect in other three models. The MICE model showed the best performance followed by E-M model. Among imputation methods, final models were not the same, in terms of composition and performance. Therefore, modern imputation methods are recommended to recover the information

Multiple imputation in survival models: applied on breast cancer data

Missing data is a common problem in cancer research. While simple methods such as complete-case [C-C] analysis are commonly employed for handling this problem, several studies have shown that these methods led to biased estimates. We aim to address the methodological issues in development of a prognostic model with missing data. Three hundred and ten breast cancer patients were enrolled. At first, patients with missing data on any of four candidate variables were omitted. Secondly, missing data were imputed 10 times. Cox regression model was fitted to the C-C and imputed data. Results were compared in terms of variables retained in the model, discrimination ability, and goodness of fit. Some variables lost their effect in complete-case analysis, due to loss in power, but reached significance level after imputation of missing data. Discrimination ability and goodness of fit of imputed data sets model was higher than that of complete-case model [C-index 76% versus 72%; Likelihood Ratio Test 51.19 versus 32.44]. Our findings showed inappropriateness of ad hoc complete-case analysis. This approach led to loss in power and imprecise estimates. Application of multiple imputation techniques to avid such problems is recommended

Dichotomisation of continuous data: review of methods, advantages, and disadvantages

In medical research, dichotomisation of continuous variables is a widespread use approach. However, it has been argued that dichotomization might be waste of information. The aim of this paper is to review the main methods to dichotomise continuous data, to address practical issues around dichotomization methods, and to investigate whether dichotomisation is always a bad idea. A total of 310 breast cancer patients were recruited. Information on 3 categorical and 1 continuous variable [age at diagnosis] was available. Missing data were imputed applying the Multivariable Imputation via Chained Equations [MICE] method. Then a minimum P-value method was applied to dichotomise the age variable. The Cox regression model was fitted to develop models in which dichotomised versus continuous version of the age variable plus other 3 variables were used. Results were compared in terms of discrimination ability, goodness of fit, and classification improvement. For the age variable, an optimal split at 47 was found. This split was close to menopause age of women in Shiraz [48] so had biological interpretability. The stability of optimal split was confirmed in bootstrap study. Model in which dichotomised version of age was used showed higher discrimination ability and goodness of fit. Furthermore, dichotomised model assigned 14% of live patients into a more appropriate risk group. Dichotomisation of continuous data is a contentious issue. We have shown that dichotomisation might improve performance of models when it has biological interpretation. More research is needed to understand situations in which dichotomisation might work

Survival models in breast cancer patients

Breast cancer is the most prevalent malignancy among Iranian women. Five and ten year survival is one of the indicators used for evaluation of the quality of care after surgery. In this study, we used several survival models to determine risk factors, survival times and life expectancies of different types of surgery. This study was performed on 310 patients who underwent surgery during a ten years period. Logistic regression and Cox regression models were used to analyze the factors leading to death. The Kaplan-Meier method [non-parametric] was used to estimate the survival rate. The log-rank test was used to compare survival in different groups. To compare life expectancy of different types of surgery, we used the actuarial life table method. Logistic regression showed that stage, grade, age and history of benign malignancy had significant relationship with death. Log-rank test showed that there was a significant difference between survival for patients with different stages, age and history of benign tumors. Cox regression model demonstrated that the variables of stage, grade, age and benign problems were the major risk factors. Actuarial life table model showed that the life expectancy for all patients was 10.03 years. This life expectancy in early stages of breast cancer for mastectomy and lumpectomy were 8.99 and 8.35 years, respectively, which was not significant. It can be concluded that the higher stage, grade, age and history of benign tumor were, the most important risk factors were correlated to mortality in breast cancer patients. This study showed that there was no significant difference between life expectancies of mastectomy and lumpectomy surgery

Anxiety, depression and anger in breast cancer patients compared with the general population in Shiraz, Southern Iran

Breast cancer and its treatment have specific challenges for women due to changes in sexuality, femininity, body image and maternal issues which might lead to associated psychological morbidity. In the present study, our aim was to measure depression, anxiety and anger in breast cancer patients. Our study comprised 178 breast cancer patients, most of whom were diagnosed for 1 to 5 years, compared with 400 other women randomly selected from the general population referred to Shiraz medical clinics for non-therapeutic reasons. We used depression, anxiety and anger subscales of Symptom Checklist-90 Revised [SCL-90R] questionnaire. Depression and anxiety were not significantly different between the two groups and the mean scores of anger were significantly lower in breast cancer patients than those in the general population. Higher depression and anger mean scores were found among younger patients. Education and tumor size correlated significantly with anxiety. Most of our patients had been diagnosed for more than 1 year and no patient had distant metastasis. Qualitative data showed stable family condition, religious tenets and social supports which are all among the reasons for our results

Primary hydatidosis of uterus and left fallopian tube

Hydatid cysts are known to occur in most organs especially in the endemic areas. However, its occurrence in uterus is extremely rare and it mostly occurs as a secondary involvement. Herein we reported on a 25-year-old woman with primary involvement of the uterus and left fallopian tube. The patient presented with a colicky abdominal pattern and fever and was admitted to the Emergency Ward

Insulin adsorbance to polyvinylchloride [PVC] surfaces of Fluid container and infusion - set

In total parenteral nutrition [TPN] solution, adsorbance of insulin to polyvinylchloride [PVC] surfaces of fluid containers and infusion-sets, decrease the amount of insulin that reaches the patients. To clarify the binding sites of insulin and to propose a solution to overcome this problem. To each of four 1000ml. PVC bottles of 5% dextrose solution, 300 microunit of insulin per each milliliter of dextrose solution were added. Each bottle was then connected to an infusion-set and the system made to run at an infusion rate of 100 ml. per hour. One milliliter samples were then collected from both the PVC bottles and infusion-sets-terminal, separately, immediately at the starting point [time zero] and 15, 30, 45 and 60 minutes thereafter. The concentrations of insulin were checked using insulin kits. At the starting point [time zero] the mean of insulin concentrations among four PVC bottles was 213.79 microunit per each milliliter of 5% dextrose solution. No significant fluctuation was noted in the concentration of insulin in the PVC bottles through 60 thminute period. However the concentration of insulin at infusion-set- terminal decreased significantly at the end of the same hour [p value=0.004]. Our results demonstrate that the adsorbance of insulin takes place at the surfaces of infusion sets. It follows therefore that increase in the primary dosage of insulin added to PVC infusion solutions and the selection of a suitable infusion set [polyethylene] seem to be beneficial for overcoming this problem

Assessment of a breast cancer screening programme in Shiraz, Islamic Republic of Iran

A breast cancer screening programme was evaluated for approximately 10,000 women aged 35 years and older. There were 67 cases of breast cancer. Highest rates of attendance were seen among younger women [35-44 years] and middle socioeconomic groups. Lowest rates were among those aged over 65 years and low socioeconomic groups. The rate of detection by self-examination was similar to that by health personnel examination. At all stages of screening, positive findings were most common among the high socioeconomic class. Attendance decreased steadily from first to last stages of serial screening. Although mammography is the most sensitive method of detection, because of its high cost we suggest establishing breast self-examination education programmes and encouraging women to self-examine

Primary tuberculosis of the male breast: a case report

A case of tuberculosis of the breast is reported in a 37-year-old man. Clinically the lesion was considered as a malignant tumor, or gynecomastia. Histological examination of the breast lesion revealed numerous caseating and non-caseating granulomas. Special [Ziel-Neelson's] stains revealed a large number of acid-fast bacilli. The patient showed satisfactory progress after anti-tuberculous therapy

Colo-pleural fistula: a rare complication of carcinoma of colon

A 72-year-old male, a known case of non-insulin-dependent diabetes mellitus, was admitted for recurrent empyema. He was found to have developed a colo-pleural fistula, secondary to a small primary carcinoma of the splenic flexure of the colon. This rare complication should be considered in the case of chronic treatment-resistant empyema

preliminary report on breast cancer screening program in Shiraz, southern Iran

Breast cancer is the commonest cancer in females in western countries. No study to determine the prevalence of this disease however has been carried out yet in Iran. We investigated 10,000 women aged 35 years and older from Shiraz, in southern Iran, to determine the prevalence of breast cancer in this region. The rough estimate of the prevalence of breast cancer was found to be 6.6 per 1000 in women aged 35 years and older [95% Confidence Interval: 2.5-10.7 per 1000]

Leiomyoma of the breast

Leiomyoma of the breast is rare. A case of leiomyoma of the breast is reported in a 36 year-old female. At first, this tumor was diagnosed as cystosarcoma phylloides cytologically. However, the histologic examination was in favor of leiomyoma, which was confirmed by immunohistological studies

Squamous cyst of the thyroid: a case report

A rare case of squamous cyst of thyroid gland is reported. Only a few cases have been reported in the literature. The exact pathogenesis of this cyst is not yet clear. Histologically this cyst can be mistaken with rare squamous cell carcinoma; and dermoid cyst, because of the presence of numerous islands of squamous epithelium. The differential diagnosis of squamous epithelium in the thyroid lesions is discussed