10/19 Presby report: geriatrics pearls

Thanks to Lindsey Haddock for an awesome and very informative SAR report on new literature in geriatrics! We discussed several important recent trials focused on older populations:


  • The “After Eighty” study: how do we treat people >80 with NSTEMI or UA?
    • this study looked at invasive intervention (PCI or CABG) vs optimum medical management in an older population with a composite endpoint: death, MI, stroke, revascularization
Screen Shot 2017-10-19 at 1.34.18 PM.png

The hazard ratio of invasive vs conservative management crosses 1 around age 90, implying that the benefit of invasive intervention fades continuously between age 80-90. The NNT for invasive intervention was 4.8. Tegn et al, Lancet 2016.

Take away: invasive intervention for UA or NSTEMI may benefit patients >80, but there may be diminishing benefit as they get older. Remember to consider other comorbidities (dementia, cancer, advanced COPD, etc) that may diminish the benefit they get from an invasive intervention!


Screen Shot 2017-10-19 at 1.52.58 PM.png

This study was a subgroup analysis of patients >75 who were part of the SPRINT trial.

  • Again, they analyzed the two groups: intervention (goal SBP <120) and conservative (goal SBP <140) within the older subgroup. They also collected data on frailty and gait speed.
  • Findings: over 3 years, intensive BP control reduced incident CV disease by 33% and mortality by 32%, with NNT = 27 to prevent one negative CV event

Screen Shot 2017-10-19 at 1.55.32 PM

Screen Shot 2017-10-19 at 1.55.23 PM

Stratified by frailty, both ‘frail’ and ‘fit’ patients seemed to benefit from intensive BP control. JAMA 2016.

The major caveat to this analysis (and to SPRINT): it excludes patients with stroke, DM, HFrEF, dementia, recent unintentional weight loss, expected survival of ❤ years, or who were nursing home residents; this obviously excludes a big proportion of the patients we see!

Take away: if an older patient is tolerating a lower blood pressure, you don’t necessarily need to de-escalate; if they have a BP >120 and don’t have other major reasons not to do add on an additional agent (falls, polypharmacy, life expectancy etc), you can consider it as it may be beneficial.


Screen Shot 2017-10-19 at 2.18.42 PM.pngThis meta-analysis suggested that in-hospital use of anti-psychotics did NOT result in improvements in:

  • short-term mortality
  • duration or severity of delirium
  • hospital or ICU length of stay

It’s important to note that many of those studies were heterogenous in terms of included patient populations, and that other studies have shown conflicting results. Many of the trials also did not evaluate for symptomatic relief, which is one of the main reasons we use antipsychotics like Haldol in the hospital

Dr. Uy pointed out that one of the reasons antipsychotics may not be as beneficial as we think is that we may just trade acute hyperactive delirium for a longer period of delirium with a longer ‘tail’, or may push them into hypoactive delirium which continues at home after discharge.


  1. Tegn et al. After Eighty Study. Lancet 2016.
  2. Williamson et al. Intensive vs Standard Blood Pressure Control and Cardiovascular Disease Outcomes in Adults Aged ≥75 Years. JAMA 2016.
  3. Neufeld K et al. Antipsychotics for Prevention and Treatment of Delirium in Hospitalized Adults: A Systematic Review and Meta-analysis. J Am Geritr Soc 2016.



10/18 Presby report: inflammatory back pain


Thanks to Dr. Lan Chen for going through her approach to lower back pain and dropping some awesome rheum pearls!

Your physical exam provides important clues to the diagnosis of AS, its extra-articular manifestations, as well as other rheumatologic conditions

  • FABER/Patrick’s test: can suggest sacroiliitis
  • Lumbar spine exam (although SI inflammation/pain is more common than L-spine pain)
  • Nails: can show pitting suggestive of psoriatic arthritis
  • Skin: psoriatic plaques, erythema nodosum, pyoderma gangrenosum, keratoderma blenorrhagicum (reactive arthritis)
  • Eyes: uveitis
  • Tongue erosions (suggestive of reactive arthritis)

Recall some key features of inflammatory back pain:

  1. Onset of back discomfort before the age of 40 years;
  2. Insidious onset lasting more than 3 months;
  3. Improvement with exercise; No improvement with rest.
  4. Pain at rest or at night
  5. Profound morning stiffness

Presence of four out of five of those features has ~70% sensitivity and specificity for ankylosing spondylitis.


Dedicated imaging of the sacroiliac joint is perhaps more useful than general spine or L-spine imaging, since sacroiliitis is more common than L-spine inflammation.

Screen Shot 2017-10-18 at 1.44.01 PM

Screen Shot 2017-10-18 at 1.47.28 PM

CT scan: normal SI joint. Note the clean, linear joint space

Screen Shot 2017-10-18 at 1.47.39 PM

CT scan: sacroiliitis. Note the shaggy, narrowed joint space 

Screen Shot 2017-10-18 at 1.47.00 PM

MRI showing marrow edema adjacent to the joint space in a patient with sacroiliitis

In early disease (first few years of symptoms), only MRI is likely to be sensitive enough to pick up SI joint changes; as the disease progresses, CT becomes more sensitive (~5 years into disease course). X-rays are really only useful in advanced disease.


Screen Shot 2017-10-18 at 2.11.52 PM.png

Zochling J, et al. Ann Rheum Dis 2006.


  • Ustekinumab and secukinumab are anti-IL23/IL17 biologics that can be used to treat AS
  • Apremilast is a PDE4 inhibitor that can also be used for AS

There are multiple causes of inflammatory back pain:

  • Primary ankylosing spondylitis
  • Psoriatic arthritis
  • Reactive arthritis
  • Enteropathic arthritis
    • Crohn’s disease
    • Ulcerative colitis
  • Juvenile ankylosing spondylitis
  • Undifferentiated spondyloarthropathies

Other rheum pearls

  • Obesity may be a risk factor for autoimmune disease
  • Gonococcal arthritis tends to more commonly affect upper extremity joints, while reactive arthritis tends to affect lower extremity joints



10/16 Presby report: MSK complications of diabetes, diabetic myonecrosis

Thanks to Samantha Parker for presenting the case of a young female type 1 diabetic who presented with bilateral thigh pain and was found to have diabetic muscle necrosis (aka myocnecrosis).

We discussed several of the MSK complications that poorly controlled diabetics are susceptible to; many of these complications are due to diabetic microvascular disease.

Approach to MSK pain in poorly controlled diabetics

  • Hand
    • Carpal tunnel
    • Dupuytren’s contracture
    • Flexor tenosynovitis
    • Diabetic sclerodactyly
    • Limited joint mobility
  • Shoulder
    • Adhesive capsulitis
    • Calcific periarthritis
    • Limited joint mobility
  • Lower limb
    • Neuropathic arthropathy (Charcot joint: foot and ankle)
    • Diabetic muscle infarction/myonecrosis (thigh and calf)
    • Osteonecrosis (knee)
    • Osteoarthritis (knee)
  • Spine
    • Diffuse idiopathic skeletal hyperostosis (DISH)

Diabetic myonecrosis

  • occurs most commonly in the thigh or calf
  • often no history of trauma, and in ~10% of cases can be bilateral
  • typically develops after 10-15 years of poorly controlled diabetes
  • CK may be moderately elevated (usually <1000), and MRI may show evidence of muscle damage and edema
  • Treatment: low dose ASA or NSAIDs, rest/analgesics, possible role for surgical excision
Screen Shot 2017-10-16 at 2.11.50 PM.png

R thigh muscle infarction. Radiopedia.

Other pearls

  • Use of warfarin and corticosteroids may predispose to calciphylaxis
  • Pyomyositis can present similarly to diabetic myonecrosis, and depending on the location of the infected muscle, can present with proximal lower extremity drainage (ex: psoas pyomyositis)
  • Complex regional pain syndrome: usually affects distal limbs characterized by pain, swelling, limited ROM, vasomotor instability, skin changes, and patchy bone demineralization; usually triggered by trauma or surgery


Cagliero E, Apruzzese W, Perlmutter GS, Nathan DM. Musculoskeletal disorders of the hand and shoulder in patients with diabetes mellitus. Am J Med 2002; 112:487.

Kapur S, Brunet JA, McKendry RJ.Diabetic muscle infarction: case report and review. The Journal of Rheumatology January 2004, 31 (1) 190-194


10/11 Presby report: PD peritonitis

We discussed an elderly woman with ESRD who presented with progressive abdominal pain, distention and fevers and was diagnosed with PD (peritoneal dialysis-associated) peritonitis, which was ultimately found to be due to TB!

What is peritoneal dialysis?

In the simplest terms, PD involves the instillation of dialysis fluid into the peritoneum with the goal of achieving solute clearance and ultrafiltration across the peritoneal membrane.

Screen Shot 2017-10-11 at 1.13.45 PM.png

Dialysis fluid ‘dwells’ in the peritoneum while solute and volume move into it; it is then removed via the peritoneal catheter at the end of the dwell

There are three specific infectious complications of PD catheters: 1) PD peritonitis, 2) tunnel infection and 3) exit site infection

The microbiology of PD peritonitis differs somewhat from spontaneous bacterial peritonitis (SBP).

Screen Shot 2017-10-10 at 11.04.20 PM.png

Microbiology of PD peritonitis. Glickman 2017

Screen Shot 2017-10-11 at 1.04.12 PM.png

  • NB: unlike SBP, where >250 PMNs or a +culture is necessary for the diagnosis, about 10% of patients with PD peritonitis present without >100 WBCs, so it’s important to have a high suspicion for it
  • It’s often a difficult decision as to whether to try to preserve the catheter or to pull it. In general, fungal, mycobacterial and pseudomonal peritonitis necessitate catheter removal.
  • You can try to treat through other forms of bacterial PD peritonitis, unless it’s a recurrent or refractory infection
  • Recurrent episodes of peritonitis can make the abdomen less hospitable to PD, and can even lead to the development of peritoneal sclerosis, which may lead to small/large bowel obstruction, constipation, malnutrition and death

This patient ultimately had her peritoneal fluid tested for adenosine deaminase (ADA) given the concern for TB. A few quick notes:

  • It can be used to distinguish between tuberculous and malignant causes of lymphocytic pleural effusions/ascites, etc.
  • Test characteristics vary somewhat between different body sites, but for tuberculous ascites, ADA (particularly values >35) has a ~95% sensitivity and 96% specificity, a positive likelihood ratio of ~16 and a NLR of 0.09!


  1. Diagnostic accuracy of adenosine deaminase for tuberculous peritonitis: a meta-analysis. Arch Med Sci 2013.

  2. Akoh JA. Peritoneal dialysis associated infections: An update on diagnosis and management. World J Nephrol 2012.

10/9 Presby report: hypertensive emergency, CSF xanthochromia, PRES

Thanks to Katie Flack for presenting a middle aged woman who woke up with a thunderclap headache and altered mental status, and was found to have hypertensive emergency. We talked about a lot of cool stuff and I tried to include most of it below!

Hypertensive emergency: classically defined as >180 systolic or >110 diastolic, or an acute rise in BP over a previously normal baseline; remember a person with lower baseline BPs can have hypertensive emergency at lower BPs!

Screen Shot 2017-10-09 at 1.38.29 PM

Varon et al 2003

Initial evaluation

  • CBC + smear (to rule out MAHA)
  • BMP (AKI)
  • Troponins (myocardial injury)
  • EKG (r/o ACS)
  • CXR (widened mediastinum, pulmonary edema)
  • Urinalysis (hematuria, proteinuria)
  • Head imaging (intracranial hemorrhage)


Most patients with hypertensive emergency are chronically hypertensive, and have a rightward shift of their vascular autoregulation curves:

Screen Shot 2017-10-09 at 1.33.06 PM.png

Varon et al 2003

This is why reducing the BP too rapidly can have lethal consequences. Important: only patients with hypertensive emergency should have their BP lowered rapidly (generally ~10% in the first hour, total ~25% in the first 24h).


  • Acute phase of ischemic stroke: keep BP <185/110 in pts who are candidate for tPA and <220/120 in those that are not
  • Aortic dissection: rapidly (within 10-20 min) lower SBP to 100-120


Briefly, agents that can be used (in the ICU) for hypertensive emergency:

  • Esmolol
  • Labetalol
  • Nicardipine
  • Nitroglycerine
  • Nitroprusside
  • Oral agents (hydralazine, nifedipine, etc)

Other pearls

  • The sensitivity of CT head for SAH declines with time: nearly 100% within 6-12h of the bleed, but it decreases to ~60% by day 5
  • In patients with a high suspicion of SAH, consider an LP even if the HCT is negative to look for RBCs or xanthrochromia in the CSF (AHA stroke guidelines, class 1B)
    • Xanthochromia can be detected 2-4h after the SAH, and is present in 90% of SAH patients within 12 hours!
  • We talked briefly about PRES: posterior reversible encephalopathy syndrome
    • Mechanism unclear, but ?related to cerebral autoregulatory dysfunction
    • Most commonly linke to hypertensive emergency or the use of immunosuppressive drugs (tacrolimus, cyclosporine > sirolimus, bevacizumab)


  1. Varon J and Marik PE. Clinical review: The Management of Hypertensive Crises. Crit Care 2003.
  2. VBGs vs ABGs: https://lifeinthefastlane.com/ccc/vbg-versus-abg/

10/4 Presby report: diagnosis of NSCLC, new targeted agents

Today we discussed the case of a middle aged man with newly diagnosed metastatic non-small cell lung cancer; this is a hot area because there have been a lot of exciting new developments in this area over the past few years– particularly the development of several new targeted agents and immune therapy.

Here’s a diagram showing the approach for a patient with newly diagnosed NSCLC:

Screen Shot 2017-10-05 at 4.39.55 PM.png

NEJM 2017


  • There are a growing list of targeted therapies based on driver mutations for NSCLC, noted in the table below
  • PD1 or PDL1 expression is important in determining the patient’s eligibility for immune therapy; 50% is the usual cut-off
    • Note that metastatic lesions may have different PD1 or PDL1 expression than the primary tumor; the reason for this is unclear, but basically means that they may still be a candidate for immune therapy even if the originally biopsied tumor had low PD expression
  • For EGFR positive patients that have disease progression on a 1st line agent like afatinib, you can get more tumor tissue and look for the T790M mutation which would make them a candidate for osimertinib (which per recent data is increasingly being used as a first line therapy for patients with EGFR exon 19 or 21 mutations)
  • It is exceedingly rare for more than one driver mutation to occur in one patient; if that happens, it may mean there are two primary cancers which would also be unusual
  • These targeted therapies don’t have the same adverse effects as standard chemo, so oncologists may have a lower threshold for giving them to patients with low performance status
Screen Shot 2017-10-04 at 9.59.31 AM

NEJM 2017


Lastly, remember that there is a trial showing the survival benefit of early palliative care in NSCLC. Obviously this finding can be extended to other types of malignancies, but at least here at Penn is mostly limited by the availability of access to palliative care.


  1. Precision Diagnosis and Treatment for Advanced Non–Small-Cell Lung Cancer. NEJM 2017.
  2. Temel J et al. Early Palliative Care for Patients with Metastatic Non-Small Cell Lung Cancer. NEJM 2010.

10/5 Presby report: possible PID

Thanks to Jake Martin and Malcolm Kearns for presenting a perplexing case of a young woman with several days of high fevers and abdominal pain, which later progressed to include transaminases in the 400s, a CK >1000, and RUQ pain–> all of which was ultimately thought to be due to pelvic inflammatory disease (although that’s not completely confirmed)!

What is pelvic inflammatory disease?

  • PID is an ascending infection that goes up from the cervix up to the uterus, fallopian tubes, ovaries and can even spread intraperitoneally, leading to liver capsule inflammation (Fitz-Hugh-Curtis syndrome)
    • PID can lead to high rates of infertility despite treatment: in one study, ~20% of women with treated PID reported infertility or ectopic pregnancy, suggesting that inflammation itself could lead to long-term damage despite adequate antimicrobial treatment

Clinical manifestations/microbiology

  • PID encompasses a broad spectrum of clinical manifestations (see table below)
    • Acute symptomatic PID: acute lower abdominal/pelvic pain, pelvic organ tenderness, possibly abnormal uterine bleeding or dyspareunia, RUQ pain if perihepatitis. Fever may not be present.
    • Subclinical PID: more indolent, with more atypical manifestations
    • Chronic PID: low grade fever, abdominal pain and weight loss over a long period, particularly associated with TB and Actinomyces (?association w/ IUDs)
Screen Shot 2017-10-05 at 2.53.41 PM.png

NEJM 2015


PID is a clinical diagnosis, requiring:

  • Pelvic organ tenderness (CMT, uterine compression tenderness on bimanual exam, adnexal pain) PLUS
  • Lower GU tract inflammation (endocervical exudate or as yellow/green mucus on swab placed gently into the cervical os (positive “swab test”); cervical friability or increased WBCs on wet mount of vaginal secretions

The presence of fever or leukocytosis can help, but are not necessary. Imaging (TVUS, CT, MRI) can also be helpful: thickened, fluid-filled tubes/oviducts with or without free pelvic fluid could represent salpingitis, or tubo-ovarian abscess.

Unfortunately this clinical diagnosis only about 60-70% sensitive, which highlights the importance of empiric treatment given the high risk of withholding antibiotics.

All patients w/ suspected PID should undergo: vaginal exam w/ wet mount of secretions (to evaluate for increased WBCs), GC testing, HIV, RPR, pregnancy test, +/- ESR/CRP

We touched on the testing characteristics of GC testing. You can test for GC using NAAT (= gold standard), culture or gram stain.

  • For women, NAAT screening obtained by vaginal swab is best; endocervical swab is fine if you’re doing a pelvic exam anyway, but not necessary to do one just for NAAT
  • NAAT is >99% sensitive and specific for urogenital gonorrhea from cervical specimens (urine is about 10% less sensitive)

Lastly, treatment options

Screen Shot 2017-10-05 at 4.11.23 PM.png

Treatment duration is usually 14 days. Interestingly, removal of an indwelling IUD does not hasten resolution (and may even worsen it)


  1. Recommendations for the Laboratory-Based Detection of Chlamydia trachomatis and Neisseria gonorrhoeae — 2014

  2. Pelvic Inflammatory Disease. NEJM 2015.