DITA markup for (hidden) indexterms vs. (visible) keywords

DITA_indexterms-vs-keywords.xml

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE reference PUBLIC "-//OASIS//DTD DITA Reference//EN" "reference.dtd">

<reference id="ID">

  <title id="Module">Module Name</title>

  <prolog>
    <metadata>
      <keywords>
        <!-- These keywords are included in index but not visible in output -->
        <indexterm>$KEYWORD1$</indexterm>
        <indexterm>$KEYWORD2$</indexterm>
        <indexterm>$KEYWORD3$</indexterm>
      </keywords>
    </metadata>
  </prolog>

  <refbody>
    <section>
      <title>Keywords</title>
      <!-- These keywords are visible in output but not included in index -->
      <keyword>$KEYWORD1$</keyword>, <keyword>$KEYWORD2$</keyword>, <keyword>$KEYWORD3$</keyword>
    </section>
  </refbody>
</reference>

For a software reference project, I need to include keyword metadata in topics so that the keywords appear both in an index and in the topic body output.

<indexterm> elements specified in the prolog metadata are included in the index but not visible in output, whereas <keyword> elements that appear in the topic body are visible in output but not included in the index.

Is there any way to avoid this redundant markup and specify the keywords once so they appear in both the index and in output?