Thursday, September 15, 2016

Liver cirrhosis mortality, alcohol consumption and tobacco consumption over a 62-year period in a high alcohol consumption country: a trend analysis


Background:

The purpose of this study is to examine the trends within liver cirrhosis mortality and its relationship to alcohol and tobacco consumption. Pulling data from several countries aids this study in an effort to improve forecasting accuracy for alcohol consumption and future mortality rates for liver cirrhosis.

Methods and Discussion:

For this paper, trends were calculated using Joinpoint regression analysis. This type of regression analysis provides the beginning and end results for a single trend. The data provided throughout the study was run through this regression analysis for the purposes of discovering increasing or decreasing trends in liver cirrhosis. Within this study, trend analysis revealed a couple key results. First, the decrease starting in 1969 among females and in 1977 among males concerning liver cirrhosis mortality, is related to decreases in alcohol consumption starting in those same years. Lastly, the same results were related to decreased in tobacco consumption starting in 1972. Figure 1 below illustrates that since the 1970’s there has been a decrease in liver cirrhosis mortality while at the same time alcohol and tobacco sales and alcohol treatment has also been on the decline.


Criticism:

Authors John Ulrich and Monika Hanke state that this approach offers less bias compared to studies using national survey data. However, this study also displayed many assumptions about what may and may not be the cause for decreased trends in liver cirrhosis. The authors did not go into detail about what actually accounted for the decreasing trend, rather it appeared that they gave their best educated guess as a result of the analysis.

Source:

Ulrich, J., & Hanke, M. (2015). Liver cirrhosis mortality, alcohol consumption and tobacco consumption over a 62-year period in a high alcohol consumption country: a trend analysis. BioMed Central Research Notes, 8, 1-7.

9 comments:

  1. It seems like trend analysis may be a useful method in analyzing the relationship between liver cirrhosis and alcohol and tobacco consumption. Perhaps, though, the study should have expanded its analysis to other variables so that it did not have to make assumptions.

    ReplyDelete
    Replies
    1. Yes I agree with you. The study should have included other variables to avoid the assumptions it made.

      Delete
  2. Charles, did this study by chance come out and delineate which countries were incorporated into the sample? When I was searching for studies to review, I noticed a series of studies along this same vein which were referencing countries who may "culturally" have a higher predisposition to both alcohol and tobacco consumption than others.

    ReplyDelete
    Replies
    1. Yes Thomas, the article was geared more toward Germany. Otherwise, it did make several references to the United States.

      Delete
  3. Chad, this is an excellent example of trend analysis. I think the main reason the researchers could only guess as to the actual cause of the decreasing trend is because correlation does not imply causation. The causal reason was likely outside the scope of their study and would be of interest for future research.

    ReplyDelete
  4. I would agree with Hank. Also you are studying something in the late 1960s and into the 1970s. All one has is the raw data on hand and not necessarily access to further knowledge that could alter assumptions made. Overall I found this piece interesting, and would like to see maybe an application of it in the modern day.

    ReplyDelete
  5. Charles, interesting article you found. Did the authors make any attempts to predict future levels of liver cirrhosis or did they just look at the available data at a given point in time and run regressions to see if they could find something to predict what they already knew happened? It seems like they already had a lot of data at their disposal. I'm not surprised deaths decreased as sales decreased. Seems pretty intuitive.
    -Eric S.

    ReplyDelete
    Replies
    1. The authors did hint toward forecasting future levels of liver cirrhosis, but they did not go into much detail on this.

      Delete
  6. This comment has been removed by the author.

    ReplyDelete