{
  "id": 170036,
  "title": "Question about data",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/170036",
  "author_name": "",
  "post_date": "2020-07-26T07:27:42.092295200Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I`m analyzing test.csv and wondering how this is possible for those two entry:\n|Patient                                             |Weeks     |FVC    |Percent    |Age    |Sex    |SmokingStatus|\n|ID00419637202311204720264  |6          |3020   |70.186855  |73     |Male   |Ex-smoker*<em>|\n|ID00422637202311677017371  |6          |1930   |76.672493  |73     |Male   |Ex-smoker</em>*|</p>\n\n<p>They have difference only on FVC and Percent (Age, Sex, Smoking are identical). And for greater FVC we have smaller Percent. I understood that for greater FVC we should have greater Percent.</p>\n\n<p>Any thoughts about this? Maybe there is another parameter which has impact on \"Percent\" which is not presented here.</p>\n\n<p>&gt; Percent- a computed field which approximates the patient's FVC as a percent of the typical FVC for a person of similar characteristics</p>\n\n<p>Or maybe it is completely artificial data, so I don`t have to worry about its integrity...</p>",
  "messages": [
    {
      "id": "945862",
      "postDate": "07/26/2020 07:27:42",
      "content": "<p>I`m analyzing test.csv and wondering how this is possible for those two entry:\n|Patient                                             |Weeks     |FVC    |Percent    |Age    |Sex    |SmokingStatus|\n|ID00419637202311204720264  |6          |3020   |70.186855  |73     |Male   |Ex-smoker*<em>|\n|ID00422637202311677017371  |6          |1930   |76.672493  |73     |Male   |Ex-smoker</em>*|</p>\n\n<p>They have difference only on FVC and Percent (Age, Sex, Smoking are identical). And for greater FVC we have smaller Percent. I understood that for greater FVC we should have greater Percent.</p>\n\n<p>Any thoughts about this? Maybe there is another parameter which has impact on \"Percent\" which is not presented here.</p>\n\n<p>&gt; Percent- a computed field which approximates the patient's FVC as a percent of the typical FVC for a person of similar characteristics</p>\n\n<p>Or maybe it is completely artificial data, so I don`t have to worry about its integrity...</p>",
      "rawMarkdown": "I`m analyzing test.csv and wondering how this is possible for those two entry:\n|Patient \t                                         |Weeks \t|FVC \t|Percent \t|Age \t|Sex \t|SmokingStatus|\n|ID00419637202311204720264 \t|6 \t        |3020 \t|70.186855 \t|73 \t|Male \t|Ex-smoker**|\n|ID00422637202311677017371 \t|6 \t        |1930 \t|76.672493 \t|73 \t|Male \t|Ex-smoker**|\n\n\n\nThey have difference only on FVC and Percent (Age, Sex, Smoking are identical). And for greater FVC we have smaller Percent. I understood that for greater FVC we should have greater Percent.\n\nAny thoughts about this? Maybe there is another parameter which has impact on \"Percent\" which is not presented here.\n\n&gt; Percent- a computed field which approximates the patient's FVC as a percent of the typical FVC for a person of similar characteristics\n\nOr maybe it is completely artificial data, so I don`t have to worry about its integrity...",
      "votes": null
    },
    {
      "id": "946885",
      "postDate": "07/26/2020 23:26:41",
      "content": "<p>A possible reason is that theyre using the <a href=\"https://www.merckmanuals.com/medical-calculators/PulmonaryPredictM_A.htm\">Adjusted Predicted Values</a>, which relies on age, height, race to predict expected FVC. For example, if the first patient is about 5 foot 10 inches tall (~178 cm) and Caucasian, this calculator would predict the FVC to be 4340 cc, which would be close to the 70% predicted. So there's other variables being compressed into percentage.</p>",
      "rawMarkdown": "A possible reason is that theyre using the [Adjusted Predicted Values](https://www.merckmanuals.com/medical-calculators/PulmonaryPredictM_A.htm), which relies on age, height, race to predict expected FVC. For example, if the first patient is about 5 foot 10 inches tall (~178 cm) and Caucasian, this calculator would predict the FVC to be 4340 cc, which would be close to the 70% predicted. So there's other variables being compressed into percentage.",
      "votes": null
    },
    {
      "id": "949084",
      "postDate": "07/28/2020 12:33:43",
      "content": "<p>Thanks, that's a useful answer for describing the Percent column. Now I wonder could we use the three column \"Age\", \"Sex\" and \"SmokingStatus\" to predict a typical FVC for a person of the characteristics(rather than Adjusted Predicted Values does).</p>",
      "rawMarkdown": "Thanks, that's a useful answer for describing the Percent column. Now I wonder could we use the three column \"Age\", \"Sex\" and \"SmokingStatus\" to predict a typical FVC for a person of the characteristics(rather than Adjusted Predicted Values does).",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 946885,
      "author_name": "jjinho",
      "author_url": "",
      "post_date": "07/26/2020 23:26:41",
      "content": "<p>A possible reason is that theyre using the <a href=\"https://www.merckmanuals.com/medical-calculators/PulmonaryPredictM_A.htm\">Adjusted Predicted Values</a>, which relies on age, height, race to predict expected FVC. For example, if the first patient is about 5 foot 10 inches tall (~178 cm) and Caucasian, this calculator would predict the FVC to be 4340 cc, which would be close to the 70% predicted. So there's other variables being compressed into percentage.</p>",
      "votes": null,
      "replies": [
        {
          "id": 949084,
          "author_name": "cathesilta",
          "author_url": "",
          "post_date": "07/28/2020 12:33:43",
          "content": "<p>Thanks, that's a useful answer for describing the Percent column. Now I wonder could we use the three column \"Age\", \"Sex\" and \"SmokingStatus\" to predict a typical FVC for a person of the characteristics(rather than Adjusted Predicted Values does).</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "945862": "I`m analyzing test.csv and wondering how this is possible for those two entry:\n|Patient \t                                         |Weeks \t|FVC \t|Percent \t|Age \t|Sex \t|SmokingStatus|\n|ID00419637202311204720264 \t|6 \t        |3020 \t|70.186855 \t|73 \t|Male \t|Ex-smoker**|\n|ID00422637202311677017371 \t|6 \t        |1930 \t|76.672493 \t|73 \t|Male \t|Ex-smoker**|\n\n\n\nThey have difference only on FVC and Percent (Age, Sex, Smoking are identical). And for greater FVC we have smaller Percent. I understood that for greater FVC we should have greater Percent.\n\nAny thoughts about this? Maybe there is another parameter which has impact on \"Percent\" which is not presented here.\n\n&gt; Percent- a computed field which approximates the patient's FVC as a percent of the typical FVC for a person of similar characteristics\n\nOr maybe it is completely artificial data, so I don`t have to worry about its integrity...",
    "946885": "A possible reason is that theyre using the [Adjusted Predicted Values](https://www.merckmanuals.com/medical-calculators/PulmonaryPredictM_A.htm), which relies on age, height, race to predict expected FVC. For example, if the first patient is about 5 foot 10 inches tall (~178 cm) and Caucasian, this calculator would predict the FVC to be 4340 cc, which would be close to the 70% predicted. So there's other variables being compressed into percentage.",
    "949084": "Thanks, that's a useful answer for describing the Percent column. Now I wonder could we use the three column \"Age\", \"Sex\" and \"SmokingStatus\" to predict a typical FVC for a person of the characteristics(rather than Adjusted Predicted Values does)."
  },
  "source": "meta"
}