Back Pain (& Neck Pain)

Fortunately this blog only deals with diagnosis of back pain, not treatment.

Low back pain is one of the only symptoms for which history and chronology don’t help very much.  The complaint is so common that I go through my same routine & thought process every time, regardless of how long-standing a patient’s pain has gone on.

HISTORY  —  The presence or absence of trauma hardly changes the differential diagnosis.  Most of the same conditions we think about with acute pain can present sub-acutely or even chronically.

The vast majority of cases of Low Back Pain are due to 3-4 conditions, diagnosed primarily by physical exam.


Target on Physical Exam

Uncommon Conditions Not to Forget

First, a few words about the uncommon causes:

**  Ankylosing Spondylitis  —  For patients with low back pain of more than several months duration, ask if symptoms are worst in the morning upon awakening, and then abate as the day goes on.  If so, perform a Modified Schober Test to measure for abnormal persistence of lordosis upon flexion.  Order a sed rate (may approach 100).  Consider a 1-view pelvic film for sacroiliac joints (NOT a “low back series”).  Send to Rheumatology if the sed rate is high, or if there’s sacroiliitis on a film.  [trivia: a double “i” in this word; “ileitis” with “e” is Crohn’s disease].

**  Epidural Compression (Cauda Equinae syndrome)  —  Worry about this possibility with acute low back pain in injection drug users (abscesses), and in patients with a history of prostate, breast, or lung cancer (metastatic infiltration).  Suspect if pain is worse when supine.  Physical exam should reveal focal tenderness to percussion of one vertebra, as it will with vertebral osteomyelitis as well.  Red Flags for potential neurological catastrophe:

  • bilateral objective motor or sensory deficits in the legs
  • bowel / bladder dysfunction
  • perineal / perianal paresthesias
  • any new focal pain in a patient at risk.

If suspicious, test for loss of anal wink reflex [have a decent index of suspicion; don’t incorporate this in your typical exam].  Need an MRI to diagnose.  Send to E.D.; paraplegia can occur suddenly.

**  Bone Cancer (Metastatic, or Multiple Myeloma)  —  Suspect in older patients if the pain is focal to a vertebra.  History may reveal weight loss, fevers, or night sweats.  In the right setting, a sedimentation rate (ESR) >50 may have up to 97% specificity for cancer as a cause of low back pain.  Plain films are usually diagnostic.

You can usually identify bone disease like cancer on physical exam by finding focal tenderness to percussion on one vertebra.  But even if I don’t elicit this, I still order plain films on older persons with new-onset low back pain without a good explanation.

**  Abdominal Pathology Radiating to the Back  —  Consider conditions such as renal stones, ectopic pregnancy, ruptured aortic aneurysm, etc., when patients at-risk present with acute back pain unaffected by movement or lumbar palpation.

**  Zoster  —  Like we saw for chest & abdomen, Zoster can cause pain anywhere.

NOW THAT WE’RE DONE WITH THE “ZEBRAS”  [click here if you never heard of “zebras”], onward to:


Diagnose these based primarily on your physical exam:

**  Bone Disease (fracture, metastasis, infection)  —  Tender to percussion of a single vertebra.  No other tenderness.  Confirm with x-ray.  These conditions, which we just discussed, actually aren’t common, but they’re the main rule-out during exam, so we think about them.

**  Herniated Disk  —  Pain radiates down a single dermatome.  Unilateral motor weakness, decreased sensation, or decrease/loss of a DTR.  Positive straight-leg raise (SLR).

**  Spinal Stenosis  —  Here history is actually key: Spinal Stenosis causes “neurogenic claudication,” i.e. pain with walking.  It primarily occurs in the elderly.  In contrast to vascular claudication, which is relieved almost immediately upon rest, spinal stenosis persists unless the person sits or lies down, & then still requires several minutes.  Pain can also be relieved by bending forward (like on a shopping cart), even if the person continues ambulating (not so with vascular claudication).

**  Muscle Strain / Ligament Sprain  —  Tender range of motion, diffuse tenderness to palpation or focal muscle tenderness (not focal to a bone), and NO FINDINGS suggestive of bone or disk disease.


A patient looks comfortable at rest, winces with movements.  Maybe there was a precipitating event, but maybe not.  Strains or sprains can occur from minor abrupt movements, with pain beginning 1-2 days later as swelling develops within tissue fibers.  Disk herniation can be abrupt or insidious.

Step 1 — Ranges of Motion: flexion, hyperextension, lateral bending, rotation.  Expect at least one to be tender.  See how disabled they are.

Step 2 — Seek Focal Bony Tenderness to Percussion:  Lightly palpate each lumbar vertebra.  If no tenderness, percuss.  Still non-tender, percuss harder.  Pound.  No focal tenderness means no bone lesion or epidural catastrophe.

Step 3 — Palpate the paralumbar muscles.  Lack of tenderness doesn’t exclude a simple strain or sprain, but if tenderness or spasm are elicited, it confirms these eventual diagnoses, & convinces the patient.

Step 4 —  Test Motor Strength (ideally with pt. supine).  Have them flex their hips against resistance while you press down on the things.  Extend knees against resistance.  Can also test hip adduction, abduction, internal & external rotation.  Ankle dorsiflexion.  Hip extension, knee flexion, and ankle plantarflexion are most accurately tested prone (against gravity).  Confession — I virtually never have a patient roll over for this (waiting-room full, time at a premium [not “time is money,” because our patients are all indigent & uninsured]).

  • THE BIG TOE !!!  —  The most important group of all for motor testing, dorsiflexion & plantarflexion.  Never forget to test the big toe!  That’s because unilateral motor weakness suggests a herniated disk, & a serious one at that.  The most common ones are L4-L5 and L5-S1, which govern the big toe.

 Possible Confounder  —  Factitious Motor Weakness, usually not frank malingering, but occurring among people so upset by and focused in their pain to be convinced that everything’s wrong.  Findings include “break-away,” and “Hoover’s Sign”.

Step 5 — Test the Deep Tendon Reflexes (patellar & Achilles), looking for unilateral diminished DTR (a sign of a herniated disk).

Step 6 — Straight-leg Raise (SLR).  With patient supine, cradle the heel & lift a relaxed straight leg.  At point of pain, pause; then abruptly dorsiflex the ankle.  Positive finding is when the latter maneuver evokes a grimace or “Ow”.  But you then have to ask, “Where did that hurt?”  Low back is positive.  Contralateral low back is strongly positive (called a “cross SRL”).  Pain felt in the posterior knee is meaningless (just indicates tight hamstrings).

Confession #2 — I often test neuro function with patients seated instead of supine, gaining valuable minutes in a busy clinic.  But when patients appear significantly disabled, I have them lie down for the exam.


If there’s focal tenderness to a single vertebra, no other tenderness, & no reason to suspect rare epidural pathology [see below], get an L-S Spine film.  Also consider an X-ray for new back pain in the elderly (cancer risk).  But these situations are most uncommon (so you won’t wind up ordering many films).

Otherwise, we’re left with Herniated Disk vs. Benign Strain / Sprain.  Look for a disk; if no findings, it’s Muscle.  So diagnose Herniated Disk if any unilateral findings:

  • Motor Weakness, especially of the lower nerve roots.
  • Diminished DTR
  • Positive Straight-Leg Raise
  • Paresthesias (subjective) in dermatomal distribution.

The latter two parameters, without weakness or reflex changes, are enough to convince me that a disk is possible.  The former 2 deficits indicate the problem is more serious.

Diagnose spinal stenosis in the elderly patient with exertional low back pain that persists despite resting, but diminishes with sitting or bending forward, even as the person continues ambulating.

Imaging for Low Back Pain????  Almost never indicated!!!!  Studies invariably reveal very low rates of serious pathology such as cancer (in the range of 0.1%).  Evidence-based clinical algorithms for low back pain testing aren’t very helpful.  Lists of “Red Flags” have high false-positive rates, and little value in primary care settings.  One study found “clinician judgment” was a useful criterion.

So, modestly speaking, what’s my “clinical judgment”?

I send patients to the E.D. for urgent MRI if I suspect epidural pathology [extremely rare], based on some combination of:

  • Perineal / Perianal paresthesias, Loss of Anal Wink
  • Bilateral leg weakness
  • Focal tenderness to percussion over a vertebra
  • Fever (objectively)
  • IV Drug User / Hx of Cancer (esp. prostate, breast, lung) / Immunocompromised

I order plain films for suspicion of fracture (pathologic or traumatic):

  • Focal tenderness to percussion over a single vertebra (with no neuro deficit)
  • Weight loss, night sweats
  • New, unusual pain in an elderly person, or patient on chronic steroids

In terms of MRIs for herniated disks or spinal stenosis, I stall as long as possible.  The problem is that they lack specificity.  Lots of people with no back pain at all have a wide variety of positive disk findings.  Almost 50% of adults over 60 may have an image positive for spinal stenosis.  Orthopedists & Neurosurgeons ignore MRI findings unless they correlate with clinical signs or symptoms of specific nerve roots.

I explain that the people who require MRIs are the surgical candidates [why do a test unless you intend to act on it?].  There are 2 treatments for herniated disks: a) Time (around 70% resolve spontaneously within 3 months); and 2) Surgery.  When patients hear this, they’re more than happy to “Give Time a Chance.”

In terms of spinal stenosis, the options are ongoing analgesia or surgery [time won’t work; physical therapy & epidural injections have never been shown to help].  If the pain lasts long enough that a patient would truly opt for the procedure, by all means request the image.  A significant or progressing neurological impairment requires surgery much sooner.

That’s it for the back.  On, very briefly, to…


Same differential as Low Back, with one big difference — there’s the spinal cord.  Cord Compression from epidural pathology can cause quadriplegia or worse [if there is such a thing].

Neck pain is usually due to a Strain or Sprain.  But there could be a Herniated Disk trapping a nerve root (radiculopathy).  Like the legs, look for unilateral motor weakness in an arm, pain / paresthesias in an upper extremity dermatome, or a unilateral diminished reflex.  There’s no such thing as a “straight arm raise” or other SLR equivalent.

Of course, percuss the vertebrae to rule out bone disease, like we discussed.

And to make sure there’s no Cord Compression, consider the legs!  Be sure there’s no motor weakness there, or hyperactive reflexes / ankle clonus / upgoing Babinski’s.  Cervical disk disease can affect nerve roots in the arms, but if something’s wrong in the legs, it points to the Cord [giving upper motor neuron findings].  Other suggestions include bowel / bladder dysfunction, perineal / perianal paresthesias, loss of Anal Wink or superficial Abdominal Reflexes.

And then there’s the very useful symptom Lhermitte’s sign — neck flexion sends an electric shock running down the body.  It’s pathognomonic of cervical cord pathology, common in Multiple Sclerosis, but in the context of “Neck Pain,” suggestive of ominous Cord Compression.

One last note about the neck — Torticollis, also called “Wryneck.”  Most common in kids under 5, it can occur in anybody due to muscle spasm.  The main imitator is a dystonic reaction to medication such as phenothiazines or metaclopramide, but also carbamazepine or phenytoin; reversible with H-1 antihistamines (e.g. Benadryl, best IV).

In my Kentucky ER, I saw 2 teenage boys from a ritzy prep school with fulminant dystonia due to some pill they’d scored, thinking it was “Valium” (was likely Haldol).  We were initially reluctant to give them the Benadryl, lest they recover & kill the dealer.  [on our shift].

That’s it for “Pain in the Neck.”  In terms of assessing pain anywhere in the body, see our posting Musculoskeletal Pain.

2 responses to “Back Pain (& Neck Pain)

  1. Thank you! Your comments about when to order an MRI are so true. I have difficulty explaining that to some of my patients.

  2. Thank you! This is so useful!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s