Rimal Mousa

University of Jordan | UJ · Department of Biopharmaceutics and Clinical Pharmacy

 I need some help. I am trying to make some assumptions about the Primary care cost for older women with primary breast cancer, since I did not get the data. I am wondering if you have any idea about the frequency of GP visit of breast cancer who are followed in the primary care?and if there is a cost for prescribing the drug in the primary care. What about other resources incurred in the primary care, i.e nurse visit? Does GP cost include the cost of prescribing the drug? Many thanks

I am trying to fit a survival model for death rate, using age as a continuous variable and oestrogen receptor (ER) status as a categorical variable. Then, I will use the fitted model to predict the death rate for patients with age 70 years being adjusted for ER status using STATA . I do not want to specify certain category within the ERstatus. However, STATA gave a coefficient for each ER status compared to the reference group rather than give one coefficient for ERstatus as the following equation and attached file.

xi: streg i.ERStatus_cat_positive Age , nohr d(weibull) nolog 

I tried to omit xi, however, STATA would treat ER status a continuous variable. In addition, if single coefficient would be derived, I do not know what is the ERstatus value (i.e. the code) that I need to multiply the coefficient with. 

Does anyone know how can I predict the death rate considering ERstatus as one group without being separated into categorical groups? Moreover, if there is a way to derive a single coefficient for ER status, how can I interpret this single coefficient, i.e. what is the ERstatus value or code number that I have to add to the coefficient to represent the overall ER status impact within the model

I am trying to fit a gamma distribution for costing data, the only information that I have are mean plus interquartile range from NHS reference costs. However, in order to fit a gamma distribution, the standard error and mean are required. Does anyone know how to use the interquartile range to estimate the gamma distribution parameters that later will be used to conduct a probabilistic sensitivity analysis for economic evaluation study

 I am currently conducting a project in older women with  breast cancer, and I need to have some estimation about the life expectancy for older women over 70 years old with breast cancer in the UK. These life expectancies would be either for each individual age over 70 years old in the UK, i.e. for 70, 71, etc. or for each specific subgroup, i.e. 70-75, 75-80, 80-85, etc.

During my search, I came across the London School of Hygiene and Tropical Medicine website where it provided some life tables from The CONCORD-2 study. However, None of these life tables were specific for breast cancer.Moreover,  the cancer life table that I downloaded from the London school website (Attachment) does not contain the number of survivor (lx) and death for each exact age (dx) that are necessary to calculate the life expectancy. I would be highly appreciated if anyone can provide me with any life table data that can help me to estimate the life expectancy  for patients over 70 years with breast cancer (either as an individual age or age subgroup)  or anyone already have a summary about these life expectancies by age.  

It is really very crucial to my research to provide me these data since I do not any access to these data.  I would cite any data you provide me with in my thesis, and I am willing to have any confidentiality agreement regarding any data you provide to me.

I am conducting cost -effectiveness analysis comparing different treatments. I am using the life-year gained, i.e. survival time as the primary effectiveness outcomes. Although the survival times follow a normal distribution, it is inappropriate to fit a normal distribution because negative values may randomly be generated from this distribution. I am not sure what is the best distribution for fitting this data to make it probabilistic?

I used the following commands to calculate the mean differences between different treatment strategies.

oneway totalcost_alltreatment RX_cat , bonferroni

However, this command did not give me the confidence intervals around the mean differences. 

 I am trying to match two groups of treatments using  Kernal and the nearest neighbor propensity score method  . I used the following command in STATA.

psmatch2 RX_cat AGE ERStatus_cat, kernel k(biweight)

psmatch2 RX_cat AGE ERStatus_cat, nn(5)

Where RX_cat stand for treatments, and ERStatus stand for estrogen receptors. 

This command gave me the propensity score for each treatment . However, I could not separate the new matched group  in a separate variable so I can analyse them separately,i.e. identifying the matched pairs with specific ID.Therefore my question is what the command the I can use to create another column or variable for the matched pairs after assigning a propensity score for them