Yeah, I think that your first sentence sums it up, Andy. There are a wide variety of results. Regarding their surgeries, I've had patients tell me, "I wish I did this years ago", and I've had others tell me, "why did I ever do this". I'm biased, but I am always an advocate of trying PT first. Why get a surgery if you don't have to. However, sometimes surgery is the route to take.
An interesting bit of research shows that if you took MRI's of a normal population sample, a good many of those people will show significant discal problems (hernations, bulges, etc). However, only a small percentage of those people showing problems on MRI will have any reports of problems from them. So in other words, some people have spines that look all jacked up on imaging, but they have no back pain. Others may have less jacked up spines, but have debilitating pain. So just take imaging with a grain of salt and focus on your symtpoms. An MRI is not the be-all-end-all when it comes to determining if surgery is needed.
Slightly off topic:
I just finished teaching a modalities course to our DPT students this semester, so I'll give those of you who are vaguely familiar with electrical treatments some more info.
///WARNING: SEVERLY NERDY CONTENT TO FOLLOW///
Most people are somewhat familiar with high rate TENS (transcutaneous electrical nerve stimulation) for pain. It's a handy, cheap, little pocket device that delivers sensory level stim to an area where you're hurting. It works by stimulating non-painful sensory nerves which in turn blocks the transmission of painful signals to the brain. Kind of like rubbing your elbow to make it feel better after you smash it on something. It really only works while you have it on, unless you're able to get a break in the muscle spasms. The good news is you can use it as much as you like.
Another option is low rate TENS. It uses a longer pulse duration and lower frequency to target motor nerves and some pain-carrying nerves, creating a rhythmic, low frequency contraction. It blocks pain through a different method: endogenous opioids. Basically, it promotes the release of endorphins and enkephalins in your body which bind with opioid receptors and can result in pain relief lasting 4-5 hours after you stop the stim. You only want to use it for 45 minutes max at a time with a break of several hours between sessions because it can lead to delayed onset muscle sorness from the muscle contraction.
Depending on the machine you have, you may be able to adjust it to acheive either one of these two effects (high or low rate TENS). For high rate TENS, you want a frequency of around 100-150 pulses per second and a pulse duration of 50-80 microseconds. Set the amplitude to a comfortable sensory level. This targets the A-beta sensory nerves I mentioned earlier to acheive "pain gating".
For low rate TENS, you want a frequency of only 2-10 pulses per second with longer pulse duration of 200-300 microseconds. Amplitude is adjusted upward until you get a visible muscle contraction. This targets motor nerves and also acheives some stim to the A-delta pain fibers which is what helps release the opioids we're seeking.
There are other options out there such as interferrential which uses crossed AC current and wave summation principles. It allows for less discomfort at the electrodes with higher amplitudes in the treatment area due to the interpherence of the two channels. However, these machines are a great deal more pricey. Many clinicians use interferential current (IFC) during treatment sessions and TENS units for home pain control.
NMES (neuromuscular electrical stimulation) is just another variation of E-stim parameters. It uses similar pulse duration to low rate TENS to acheive muscle action, but you use a higher frequency (above 35 pps) to acheive a sustained muscle contraction by stimulating the motor nerve of a particular muscle. It's uses in pain relief are pretty limited unless you just want to fatigue a muscle to reduce spasm. Russian stim is another type of muscle stim you may have heard of, but it is entirely different in parameters from NMES.
So basically, low rate TENS, high rate TENS, and NMES are all the same type of stim (pulsed, bi-phasic current) with different pulse durations, frequencies, and amplitudes. IFC, on the other hand is a different type of stim all together (interphering AC current). Where you place the electrodes depends on the goal and the type of stim being used.
We can talk about burst TENS, which tries to combine the effects of high and low rate TENS, but then we're just being nerdy.
//end nerdy content

//
--nathan