There is some low hanging fruit when it comes to accessible data visualisation: without going into technical details and the intricacies of accessible HTML, respecting some basic rules will improve the visualisation reading experience of impaired people to a great extent.
The first is to ensure enough colour contrast between text and geometrical elements and and their background. Avoid light colours for text on a white background, and certainly check the contrast ratio when using a non-white background. Also make sure that graphical elements like lines are thick enough to be clearly perceived on all kinds of screens and by people with visual impairments.
Use colour blind safe colour palettescolour blind safe colour palettes. Avoid red-green-orange-brown palettes, and make sure to test your colours in a colour blindness simulator. When the colours in a visualisation are carrying meaning and information, consider adding an additional data encoding, so that you are not relying on colour alone to make the visualisation work.
Add an understadable title to your visualisation that summarises the charts main point. Reenforce the message in the design of your visualisation with things like highlighting and annotations to make it as easy as possible for your audience to understand why you are showing them the visualisation.
When a visualisation is published online, make sure to add an alternative text. Describe the chart type and the kind of data it is showing. If the visualisation is using axes, add their titles also to the alt text, and make sure to also add the main point of the visualisation. The use of the alt text is very widespread, and its use not only benefits people using assistive technology like screen readers. It also helps the search engine optimisation of your web pages and visualisations.
When you publish a lesser known chart type, make sure to guide the viewer how to read it very clearly. Label axes in an unambiguous way and with simple language, and always add the units numbers are expressed in.
When a visualisation tool has functionalities to make the output visualisations more accessible, use them. Datawrapper, for example, has built in testing for colour blindness, a preview mode for different screen sizes, responsive visualisations and and you can add a alternative description that will be read by screen readers. Tools like this make it quite easy to make your visualisations more accessible, so you should put in the little effort to use these functionalities.