Change search
ReferencesLink to record
Permanent link

Direct link
Reasons for Missing Data of Risk Categorisation in NPCR
Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Mathematics, Applied Mathematics and Statistics.
2015 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

The risk of prostate cancer (PCa) can be described using ve risk categories based on clinicalassessment involving the risk of cancer and metastasis level. In some cases the information neededto calculate the risk is not registered. The aim is to assess potential dierences in properties likeage, treatment, comorbidity and survival, between men with a dened risk stage categorisation ofprostate cancer compared to men lacking information to calculate the risk stage category. Themain measures involved in the risk stage assessment are Gleason score, prostate specic antigen(PSA) value and T-stage. Men missing data for risk stage categorisation may be lacking one of thethree or a combination of at least two out of the three. Subgroups of these men will be analysedin a similar way, in order to understand the reasons to why they are missing data for risk stagecategorization.Statistical analysis involves univariate and multivariate logistic regression, along with survivaland competing risk analysis. Data will be presented in tables, gures and forest plots, includingodds ratios, 95% condence intervals and p-values.According to the study, men missing data of risk stage categorization were about 2.6% of allmen in the data base, and had most likely a low risk PCa. They had higher comorbidity levelsbut the overall probability of death was the same compared to other men. In addition, they hadsignicantly lower proportion of death by PCa and experienced a large proportion of death byother cancer, which concurs with the previous conclusions about comorbidity and low risk PCa,indicating that they had another disease, possibly cancer, that required more attention.Considering only men missing data for risk categorisation, a large proportion were missingPSA level (58.3%), and these men had higher comorbidity, were older, and had a large proportionof death by other cancer. Surprisingly, men missing Gleason level (20.3%) had increased oddsratios for lower comorbidity levels, were younger at time of diagnosis, and had a higher survivalprobability in general. Unexpectedly, men missing T-stage (32%) were more likely to being treatedby Radio Theraphy (RT), were less likely to attend university hospitals and more likely to attendprivate physicians. Men missing a combination of at least two out of three of Gleason, PSA, T-stage(19.6%) had higher comorbidity levels and were more likely to be treated by RT, less likely to attenduniversity hospitals, had a large proportion of death by other cancer, and a larger proportion ofdeath closer to the time of diagnosis.Lastly, there were some indications of variations of the proportions of missing data of riskstage categorisation when dividing it into the subgroups mentioned above, and viewed over year ofdiagnosis. There was an increase in missing data of risk stage categorization around 2006 and anexplanation of this could be the change of IT-system for registration, leaving a general increase ofmissing data behind it, perhaps due to a looser control during the transition and unfamiliarity ofthe new system.The main conclusion was that the reasons for missing data of risk stage categorisation are mostlikely high comorbidity levels, probably including another cancer in combination with a low riskPCa. It was most common to be missing data of risk stage categorisation due to missing PSAlevel, and those men had high comorbidity and were older. Surprisingly, private physicians and/ortreatment by RT were more likely to be missing T-stage, and younger men with low comorbiditywere more likely to be missing Gleason score.

Place, publisher, year, edition, pages
2015. , 48 p.
U.U.D.M. project report, 2015:4
National Category
Probability Theory and Statistics
URN: urn:nbn:se:uu:diva-253437OAI: diva2:814569
Educational program
Bachelor Programme in Mathematics
Available from: 2015-05-27 Created: 2015-05-27 Last updated: 2016-10-19Bibliographically approved

Open Access in DiVA

fulltext(868 kB)42 downloads
File information
File name FULLTEXT03.pdfFile size 868 kBChecksum SHA-512
Type fulltextMimetype application/pdf

By organisation
Applied Mathematics and Statistics
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
Total: 42 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 558 hits
ReferencesLink to record
Permanent link

Direct link