• 
• 

Three Part Question

In [patients with known HIV] can the [absolute lymphocyte count] accurately [identify those with a CD4+ < 200]?

Clinical Scenario

A 37-year old patient who has never been to your hospital presents for shortness of breath. He reports a history of HIV, but is not currently on treatment and does not know his last CD4+ count. His oxygen saturation is 94% on room air, and lung sounds are distant. His CXR shows possible interstitial markings in the right middle lobe, his LDH is 240, absolute lymphocyte count (ALC) is 2200, and he has a normal A-a gradient. In addition to covering for community acquired pneumonia, should TMP/SMX be started in the ED?

Search Strategy

Pubmed via Medline was searched as follows:
CD4[All Fields] AND "absolute lymphocyte"[All Fields]; CD4[All Fields] AND "absolute lymphocyte"[All Fields]; "total lymphocyte count"[All Fields] AND (estimated[All Fields] AND CD4[All Fields]). A Web of Science™ Cited Reference Search was performed for all three articles listed in the table.

Search Outcome

Three relevant papers were found.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Shapiro et al. 1998 USA	802 "consecutive samples" from outpatient, inpatient, and emergency department patients who had a CD4 and ALC over an 11-month period at one institution	Retrospective case series	Test characteristics (Sensitivity, specificity, positive and negative predictive values, likelihood ratios)	An ALC < 2000 was highly sensitive (97%) for CD4 <200, an ALC < 1000 was highly specific (96%)	No confirmation of HIV+ status and multiple different patient settings severely limits this study's applicability; unclear how many unique patients were studied
Napoli et al. 2011 USA	866 unique HIV+ patients admitted from the emergency department at one academic medical center, over an 8-year period and with a CD4 drawn within 24h of admission	Retrospective case series	Test characteristics (Sensitivity, specificity, positive and negative predictive values, likelihood ratios)	An ALC < 1700 was 95% sensitive for CD4 <200, an ALC < 950 was highly specific (93%)	The prevalence of CD4 <200 was 84.6% in this study, limiting its applicability; single-center, retrospective design
Napoli et al. 2013 USA	686 HIV+ patients admitted from the ED and with a discharge diagnosis of pneumonia at three centers over a five-year period, with a CD4 drawn within 24h of admission	Retrospective case series	Test characteristics (Sensitivity, specificity, positive and negative predictive values, likelihood ratios)	An ALC < 1700 had a sensitivity of 84% and specificity of 55% for a CD4 < 200; Empiric treatment for PCP pneumonia was only begun in 39% (62/159) patients	Also a high prevalence of AIDS (64%); retrospective design

Comment(s)

The evidence for using ALC as a surrogate for CD4+ count in HIV+ patients is quite limited. Although a calculator based on Shapiro’s 1998 paper1 can be found on MDCalc.com, it lacks validity, as none of the patients had confirmed HIV. While Napoli’s initial study demonstrated that a cutoff ALC of 1700 was highly sensitive (95%), in his follow-up study, a substantially higher ALC (2500) would be needed to achieve the same sensitivity. Other test characteristics of Napoli’s studies are skewed by the high prevalence of AIDS patients (65-85%), which significantly limit their generalizability. Further limitations to estimating CD4+ include the temporary suppression of CD4+ and ALC during acute illness and the effect of other factors (diurnal cycles, antiviral treatment, BMI, nutritional status, comorbidities) on CD4+ count.

Editor Comment

KMJ

Clinical Bottom Line

In summary, ALC cannot be recommended as a surrogate of CD4+ in HIV+ emergency department patients.

References

1. Shapiro NI, Karras DJ, Leech SH, Heilpern KL. Absolute lymphocyte count as a predictor of CD4 count. Ann Emerg Med 1998;32(3 Pt 1):323–8.
2. Napoli AM, Fischer CM, Pines JM, Soe-lin H, Goyal M, Milzman D. Off J S Absolute lymphocyte count in the emergency department predicts a low CD4 count in admitted HIV-positive patients. Acad Emerg Med 2011; 18(4):385–9
3. Napoli AM, Maughan B, Murray R, Maloy K, Milzman D. Use of the relationship between absolute lymphocyte count and CD4 count to improve earlier consideration of pneumocystis pneumonia in HIV-positive emergency department patients with pneumonia Journal of Emergency Medicine 2013: 44(1):28–35