3 of 3 problems here are unscored. That is not modesty. A problem with no citable figure for its burden cannot be ranked against anything, so finding one is the most useful thing anyone can do on this site.
The five components
A score exists only when all five are present and none is contested. Anything else is unscored, and the page says which ones are missing by name. Each needs a number, a unit, a year, and a source.
How much life the problem destroys in a year, counting death and disability together so that a disease which cripples and one which kills can be compared at all.
Usual unit: life-years lost per year
Good source: A national or global burden-of-disease study that states the figure for a named year.
Missing on 3 of 3 problems.
How many people the problem affects, whether or not it kills them. Reach and burden diverge sharply, and a problem can be enormous on one and small on the other.
Usual unit: people
Good source: A prevalence or incidence estimate with its year and its population base stated.
Missing on 0 of 3 problems.
Whether it is getting better or worse, and how fast. A problem improving quickly is a lower priority than the same problem getting worse, at equal burden.
Usual unit: per cent change per decade, signed
Good source: Two comparable measurements at least five years apart, from the same series rather than two different ones.
Missing on 3 of 3 problems.
How much effort already goes in, relative to the scale of the problem. This is the component most often missing, and it is the one that most changes a ranking.
Usual unit: 0 well funded, 100 ignored
Good source: Total annual funding from a published tally, set against burden. How this is measured is still an open question on this site.
Missing on 3 of 3 problems.
Whether more effort would actually help. A problem nobody knows how to solve and one that is solved but unfunded need different things.
Usual unit: 0 intractable, 100 solved by spending
Good source: Evidence that a named intervention works at scale, or a stated reason why the bottleneck is not money.
Missing on 3 of 3 problems.
When sources disagree
A contested component is excluded from the score, never averaged. Averaging two figures that disagree produces a third that nobody published and nobody will defend. The problem returns to unscored and the disagreement is shown, which is the honest state.