Thanks for the question! That's a good one and it comes up a fair amount. The answer is a bit tricky (ha welcome to the LSAT!), as there tends to be a difference in how causality works on the LSAT in theory, and how it tends to work in practice.
What I mean is that the example you give would imply in LSAT-land that a broken leg is a reliable outcome and the cause would always produce it, and the author assumes those two are thus always connected. But are they believed to be exclusive to the point of ruling out everything else (like a broken arm, too)? Possibly, but it's hard to know the author's full feelings on the matter as the LSAT isn't constructed to expose or examine that "extra."
That is, there's no real test scenario where an additional outcome would make a difference, good or bad: as long as something else is in addition to the leg break (and doesn't contradict either it or something else the author said), then noting that an arm was broken too
wouldn't weaken, strengthen, describe a flaw, etc. It doesn't matter, so wouldn't/couldn't be tested in a meaningful, practical sense. That's why I note that in practice it's hard to fully determine the entirety of an LSAT author's causal assumptions. The good news is that that also means they don't matter
When we say that a cause and effect exist in a vacuum we're really talking about the relationship between them, where the author presumes nothing else could have led to the described effect, and the cause as given always will. They're inextricably interconnected, to borrow Stephen Hawking's lovely phrasing (regarding the relationship between time and space, but applicable here). Authors on the LSAT feel that the linkage is unbreakable, and it's that linkage that serves as the focal point when you encounter causal reasoning.
That's also exactly why one without the other (cause without effect or vice versa), or a competing cause, are so damaging to causal arguments. They disrupt the intimacy, the constancy, assumed by the author. That's the key. Things happening in addition to the relationship, "in a separate vacuum" so to speak, are probably allowed, but irrelevant and thus not really tested.
I hope that helps!