Breathtaking mortality: bacterial pneumonia and HIV

This article originally appeared in HIV Treatment Update, a newsletter published by NAM between 1992 and 2013.
This article is more than 13 years old. Click here for more recent articles on this topic

Ever stopped to think about the top reason people with HIV end up in intensive care? Heart attack? Liver problems? Accidents? No, pneumonia. David McLay investigates an illness that scores high on the morbidity and mortality charts but gets relatively little attention.

‘The old man’s friend’, they call pneumonia, because it’s often a relatively painless coup de grâce at the end of a long life. But this image of the lung disease belies the fact that it can strike people at any age, and kill quickly if not recognised and treated. The good news is that there are things you can do to reduce the risk and treat it.

Although HIV-negative people of all ages can get serious or even fatal pneumonia too, as seen most recently with pneumonia deaths caused by swine flu, it is disproportionately more common in people with HIV.



Any lung infection that causes inflammation. The infecting organism may be bacteria (such as Streptococcus pneumoniae), a virus (such as influenza), a fungus (such as Pneumocystis pneumonia or PCP) or something else. The disease is sometimes characterised by where the infection was acquired: in the community, in hospital or in a nursing home.

Pneumocystis carinii pneumonia (PCP)

Pneumocystis carinii pneumonia is a form of pneumonia that is an AIDS defining illness.


Antibiotics, also known as antibacterials, are medications that destroy or slow down the growth of bacteria. They are used to treat diseases caused by bacteria.

pneumococcal disease

Disease caused by the bacterial infection Streptococcus pneumoniae. In most people, it causes relatively minor health problems (called ‘non-invasive’ infections) such as bronchitis, sinusitis (sinus inflammation) and middle-ear infections. It can also cause serious pneumococcal diseases including severe bacterial pneumonia, sepsis (blood poisoning) or meningitis (inflammation of the brain lining).


A group of organisms, including the yeasts which cause candidiasis and cryptococcosis.

Lower respiratory tract infections are, in people with HIV, the most common cause of hospitalisation in an intensive care unit (ICU), according to a 2007 report from University College Hospital (UCH), London.1 Nearly half of people with HIV admitted to ICUs (48%) had a pulmonary (lung) infection, with Pneumocystis pneumonia (PCP) and bacterial pneumonia being the diagnosis in 80% of them. A 2009 US study reported about 40% of HIV-positive people in intensive care were admitted with respiratory failure, including pneumonia and other lung conditions such as emphysema.2

While PCP has a storied history in the world of HIV (see Whatever happened to PCP? below), other pneumonias are not the health conditions people with HIV hear about regularly when talking to their doctors or perusing HIV news.

Dr Marc Lipman, consultant in HIV and respiratory medicine at the Royal Free hospital, London, says that people with HIV – including those on treatment – should still be aware of respiratory infections, particularly bacterial pneumonia.

With these facts in mind, let’s look at what we know about HIV and bacterial pneumonia, and what you can do to avoid it.

Defining the condition

Pneumonia is one of those ailments that are so common we often don’t bother to think much about them. Simply put, pneumonia is an infection of the lungs that causes inflammation. The infecting organism can be a bacterium, a virus, a fungus or a parasite. Streptococcus pneumoniae (pneumococcus) is the most talked-about culprit, but two other infections might be familiar to HIV-positive folk: PCP and histoplasmosis, two fungal infections common among people with severely weakened immune systems. PCP was almost unheard of until the start of the HIV epidemic in the 1980s. Fortunately, both are on the decline, thanks to effective therapies.

Because so many different microbes can cause pneumonia, the disease is usually characterised by where infection was acquired in the community, in hospital or in a nursing home. These groupings also dictate the treatment, because even if the microbe is not identified, doctors have found that cases from these different settings respond to different management.

Link to HIV

Over the course of the HIV epidemic, rates of community-acquired pneumonia have been five- to tenfold higher than in HIV-negative people of the same age, and bacterial pneumonia, 35 to 50 times higher.3,4,5,6 A Danish national cohort study of hospitalisation for pneumonia among people with HIV from 1995 to 2007 found that, while the overall rate of pneumonia decreased during that time, the rate in 2005-06 was still 6.3 times higher than among HIV-negative people.7

Bacterial pneumonia has been linked to several HIV-related factors, including CD4 counts, having an uncontrolled viral load, and not being on antiretroviral therapy. HIV-positive people with a weak immune system, particularly a CD4 count below 200 cells/mm3, are at greatest risk.  At these low CD4 counts, the risk of complications, particularly bacteraemia – the presence of the bacteria in the blood – increases dramatically. Some experts suggest that a case of pneumococcal bacteraemia is a sign that someone should be tested for HIV.


Taking all cases of pneumonia together, including the AIDS-related PCP, the Danish national cohort study found that the risk of subsequent death amongst people with HIV was almost three times higher following hospitalisation for pneumonia, compared to not having pneumonia, and lasted for more than one year,8 while a study in 2000 found that 28% of all pneumonia deaths in British young adults were among people with HIV.9

Excluding PCP, while bacterial pneumonia is more common among people with HIV, and often fatal (an Italian team reported a death rate of 26% among people admitted to ICU from 1988 to 200210), it is not clear whether HIV-positive people who acquire pneumonia are more likely to die from it than their HIV-negative counterparts.

An early case-control study during the late 1990s involving 174 people with HIV in Alberta, Canada, measured the length of hospital stay and one-year mortality in people with HIV and found a three-to-ten-times-higher immediate risk of death compared with HIV-negative people.11 However, more recent studies are suggesting that recovery is comparable with that experienced by HIV-negative people. In the UCH study, 77% of patients were eventually discharged from ICU fully recovered, compared with 74% for general medical patients.12 Being on or off HIV treatment didn’t seem to make a difference, with both groups having discharge rates near the overall rate. In a 2005 report on the international Community-Acquired Pneumonia Organization (CAPO) cohort study, the length of hospital stay and mortality rates due to pneumonia were similar in 58 HIV-positive and 174 HIV-negative people.13

Antiretroviral treatment

The introduction of effective HIV treatment has decreased rates of many infections, including pneumonia. The Danish cohort saw a significant drop in hospitalisation with the introduction of combination therapy in 1996, from 50.6 admissions per 1000 person-years in 1995-96 to 18.3 in 1999-2000.

The SMART study, which compared continuous therapy with CD4 count-guided treatment breaks, provided evidence that the effects of treatment may go beyond simple immune system strengthening. Intermittent rather than continuous therapy was associated with a 50% increase in risk of bacterial pneumonia, and continuous therapy reduced the risk of pneumonia even in those with CD4 counts above 500 cells/mm3.14

However, evidence suggests that even people with stronger immune systems may still be at risk. In the Danish cohort, the rate of hospitalisation for HIV-positive people with CD4 cell counts above 500 cells/mm3 was 5.9 times more than HIV-negative people.15

Dr Lipman explains that even in people on treatment with high CD4 counts, “There are still defects in the mucosal immunity that stop people from getting infections when they breathe in a bug.”

Risk factors unrelated to HIV

Beyond HIV infection, studies have revealed a number of risk factors for pneumonia, including:

  • Smoking (cigarettes and cannabis16)
  • Excessive alcohol drinking
  • Injecting drug use
  • Having another health condition, such as cardiovascular, pulmonary or renal disease, diabetes or cirrhosis of the liver
  • Malnutrition
  • Low socioeconomic status.

Smoking features high on this list of risks. It’s a tough road to quitting smoking (see Off the hook in issue 193 of HTU), but there are many good reasons to travel it. The pulmonary risks associated with smoking among people with HIV also include lung cancer, bronchitis and chronic obstructive pulmonary disease.


Two vaccines exist against common strains of pneumococcus, the main cause of bacterial pneumonia. They are commonly administered to children, but are sometimes recommended for adults at risk of pneumococcal disease. Both UK and US guidelines recommend vaccination with the older 23-valent polysaccharide vaccine for all people with HIV. However, there is conflicting evidence whether the vaccine provides any additional protection against pneumonia, particularly in people with low CD4 counts, who should be most likely to benefit from vaccines, but who are also least likely to produce an adequate immune response.

Results from the only randomised, double-blind, placebo-controlled trial of the 23-valent vaccine, conducted in Uganda during the early 1990s before the advent of ART, showed no protective effect. In fact, vaccination was associated with an increased risk of pneumonia, but a decreased risk of death.17,18

However, more recent studies have shown a protective effect. In a Spanish study, HIV-positive people on HIV treatment with a CD4 count above 200 cells/mm3 and people not on treatment with a count over 350 cells/mm3 had an immune response to the vaccine similar to HIV-negative people.19 In a retrospective case-control study by the same Spanish team, vaccination with the 23-valent vaccine decreased the risk of pneumococcal disease by 56%, though being on treatment reduced the risk even more, by 77%.20

While vaccination may be recommended, is it widespread? Not according to Dr Lipman, at least in his experience. Given the conflicting study results and the unproven benefit, doctors may be hesitant to vaccinate because it might give a false sense of security. Moreover, the vaccine is expensive and not covered by most public health systems.

However, a related vaccine that is routine for many people with HIV - the annual flu vaccine (which in 2010-11 included protection against swine flu) - may actually provide some indirect protection from bacterial pneumonia (pneumonia can be a complication of flu). In one Japanese study, vaccinating people with HIV for seasonal flu reduced the number of people who caught flu by 71%.21

Diagnosis and treatment

Pneumonia is diagnosed through signs and symptoms and, sometimes, a chest X-ray. In HIV-positive people with a strong immune system, signs and symptoms are similar to those seen in HIV-negative people: fever, trouble breathing, cough that produces green or yellow sputum, and possibly chest pain. These appear over the course of three to five days and persist. Any chest cold that seems to linger should be checked out by a doctor.

HIV-positive people with low CD4 counts may present with more subtle symptoms, that can be easier to miss. These include a low-grade fever, dry cough, rapid breathing and even weight loss or confusion. Diagnosis in HIV-positive people with liver disease may require special care; a recent Italian study found HIV-positive people with liver cirrhosis presented with milder symptoms, but a higher risk of death.22

Sputum and blood can be cultured in an attempt to identify the bacterium causing the infection, though successful identification is made in only about one-third of cases. As with HIV-negative people, the most common bacteria are Streptococcus pneumoniae and Gram-negative bacteria such as Haemophilus influenzae, Pseudomonas sp and Legionella sp.23

The good news is that, according to Dr Lipman, “Most bacterial pneumonias will respond as they would do in a non-HIV infected population.” Treatment involves using an antibiotic from the penicillin or related, second-generation cephalosporin family, or one from the macrolide family such as erythromycin. The antibiotic is usually prescribed without trying to identify the infecting microbe; the antibiotic can often clear the infection before test results are ready.

More important is whether people should be admitted to hospital for treatment. Doctors can use an algorithm such as the PSI (pneumonia severity index) or CURB-65 – which include basic demographic and vital signs and a short mental test – to evaluate the severity of a case. Severe cases are treated in hospital while low-risk cases can be treated at home. Neither the algorithms nor pneumonia treatment guidelines, however, have been validated for HIV-positive people. A recent review by Italian researchers suggested that PSI could be used in people with CD4 counts above 200 cells/mm3, while people with lower CD4 counts should be admitted to hospital.24

Treatment can be more complicated for people with lower CD4 counts, particularly below 100 cells/ mm3, says Dr Lipman, because they are more likely to get more exotic bugs that don’t respond to standard treatment. Also, he adds, in the early days of the HIV epidemic it was common for people to have infections with more than one organism.

It’s easy to see how bacterial pneumonia could become the forgotten sibling - plain and unassuming. Dr Lipman reminds us: “Bacterial pneumonia and other respiratory tract infections were always there. They just didn’t necessarily get the headlines that TB [tuberculosis] and Pneumocystis did.”

Effective treatments and generally good prognosis are all the more reason to pay a bit of attention to this forgotten sibling: a little prevention and early treatment can prevent an unwanted visit to ICU.

The most important thing to be aware of is that pneumonia is a common condition that can be more common in people with HIV, even in people on treatment. It is yet another good reason to give up smoking, and to disclose your HIV status to your GP, so that they know you may need antibiotics or a chest X-ray sooner than someone without HIV.

Whatever happened to PCP?

One of the things that happened to PCP is that it changed its name and should now properly be called PJP. Pneumocystis pneumonia is caused by the fungus Pneumocystis jirovecii, with the carinii species now only infecting other animals. This ubiquitous fungus is probably transmitted through airborne routes early in life; two-thirds of healthy children have antibody to P. jirovecii by the age of four. Colonisation of the respiratory tract of healthy people generally causes no ill effects. Infections appear with severe suppression of the immune system; about 90% of cases occur in people with CD4 counts below 200 cells/mm3.

PCP is seared into the collective memory of the HIV community in the resource-rich world because it was such a common and fatal condition early in the epidemic. According to the US Centers for Disease Control and Prevention, PCP occurred in 70 to 80% of people with AIDS, and even with treatment, the disease led to death in 20 to 40% of cases.25 Rates of PCP dropped – to about 2 to 3 cases per 100 people a year among people with AIDS in Western Europe and the US – with the use of effective treatments, prophylaxis and with the advent of effective HIV treatment.

The antifungal agent cotrimoxazole (Septrin or Bactrim) is the most commonly used agent for both prophylaxis and first-line treatment. Prophylaxis is started when CD4 counts fall below 200 or 250 cells/mm3. People who have already had a case of PCP often continue taking prophylaxis after infection has cleared. For people who see sustained increases in their CD4 counts to more than 250 cells/ mm3 after starting treatment, prophylaxis can be safely stopped.

Certain people are still more at risk of PCP. Dr Lipman says that PCP “generally occurs in people who present very late; in other words, who don’t know their HIV status or aren’t doing anything about it. Or they have low blood CD4 counts and have taken a regimen of prophylaxis that is ineffective or has become ineffective because they did not take it.”

My brush with pneumonia

Brian West from the European AIDS Treatment Group, formerly of HIV Scotland, recalls a summer shiver that turned into Legionnaire’s disease, one of the deadliest forms of pneumonia.

I’ve been living with HIV for over 26 years now, and thought I knew which illnesses to keep an eye out for. Pneumonia, however, was not on my list before last year.

The summer started well with a nice holiday in Spain, and four days later I got on a plane to go to the International AIDS Conference in Vienna. I started to feel a bit feverish waiting for the connecting flight at Zurich, but ignored it.

I went to bed that night and the next day things were worse. I struggled through a few meetings, but people looked at me strangely as I was shivering on a warm day. Two ex-nurse colleagues suggested going to see a doctor. I said that if I felt the same the next day I would.

Next day I woke up with a raging temperature, a headache and shortness of breath. I phoned up one of my colleagues who said she knew where to get the best treatment. Within an hour I was in a taxi going to the main hospital in Vienna. I was incredibly lucky, because having good HIV contacts meant I was taken straight to the infectious diseases unit. The senior HIV physician sent me for an X-ray and immediately diagnosed pneumonia. He then asked for sputum samples to send to the central laboratory, to see if they could isolate the type.

They told me it was legionella, or Legionnaires’ disease. I was told that it takes four to twelve days to incubate and so I would have probably caught it in Spain (later confirmed – beware of infected shower heads!). I was very lucky to have been given a precise diagnosis only five hours after I walked into the hospital - highly unusual, I believe.

The treatment was bucket loads of intravenous antibiotics over a six-day period in hospital – after 26 years of living with HIV, the longest hospital stay I have ever had. It did not end there. I had a further eight weeks of oral antibiotics (crap dosing schedules, four times a day, six hours apart - hasn’t HIV moved on?).

The recovery period was a shock. When the doctor said it would be two to three months, I thought, well maybe other people, but not me, I am fit! He was right, of course. For the first few weeks I struggled to go up a flight of stairs. I could not walk along the road and talk at the same time. This is, I discovered, multi-tasking for weakened lungs. It took me three months to feel better. Prior to this illness I had been going to the gym three times a week. I had to restart my gym regime slowly, and I can honestly say it was six months before I felt fully recovered.

This illness floored me completely, and I know I was very lucky, because I had a really quick diagnosis and the best possible care. Legionella pneumonia kills a lot of people, and I was lucky not to be one of them. I didn’t realise it was something I had to be aware of. Now I am. It’s funny how people tell you to run the shower for five minutes when you are lying attached to a drip in a hospital bed… A bit late really!

  1. Dickson SJ et al. Survival of HIV-infected patients in the intensive care unit in the era of highly active antiretroviral therapy. Thorax 62:964-968, 2007.
  2. Powell K et al. Survival for patients with HIV admitted to the ICU continues to improve in the current era of combination antiretroviral therapy. Chest 135:11-17, 2009.
  3. Nuorti JP et al. Epidemiologic relation between HIV and invasive pneumococcal disease in San Francisco County, California. Ann Intern Med 132:182–190, 2000.
  4. Feikin DR et al. Global strategies to prevent bacterial pneumonia in adults with HIV disease. Lancet Infect Dis 4:445-455, 2004.
  5. Heffernan RT et al. Declining incidence of invasive Streptococcus pneumoniae infections among persons with AIDS in an era of highly active antiretroviral therapy. J Infect Dis 191:2038-2045, 2005.
  6. Hirschtick RE et al. Bacterial pneumonia in persons infected with the human immunodeficiency virus. Pulmonary Complications of HIV Infection Study Group. N Engl J Med 333:845-851, 1995.
  7. Sogaard OS et al. Hospitalization for pneumonia among individuals with and without HIV infection, 1995-2007: a Danish population-based, nationwide cohort study. Clin Infect Dis 47:1345-1353, 2008.
  8. Sogaard OS et al. Mortality after hospitalization for pneumonia among individuals with HIV, 1995–2008: a Danish cohort study. PLoS ONE 4: e7022, 2009 doi:10.1371/journal.pone.0007022.
  9. Simpson JC et al. A national confidential enquiry into community acquired pneumonia deaths in young adults in England and Wales. Thorax 55:1040-1045, 2000.
  10. Franzetti F et al. Nosocomial bacterial pneumonia in HIV-infected patients: risk factors for adverse outcome and implications for rational empiric antibiotic therapy. Infection 34:9-16, 2006.
  11. Johnson DH et al. Hospitalization for community-acquired pneumonia in Alberta patients with human immunodeficiency virus infection: A case control study. Canadian Respiratory Journal 10(5):265-270. 2003.
  12. Dickson, op. cit.
  13. Christensen D et al. HIV infection does not influence clinical outcomes in hospitalized patients with bacterial community-acquired pneumonia: results from the CAPO international cohort study. Clin Infect Dis 41:554-556, 2005.
  14. Gordin RM et al. Pneumonia in HIV-infected persons: increased risk with cigarette smoking and treatment interruption. Am J Respir Crit Care Med 178:630636, 2008.
  15. Sogaard, 2008 op. cit.
  16. Caiffa WT et al. Drug smoking, Pneumocystis carinii pneumonia, and immunosuppression increase risk of bacterial pneumonia in human immunodeficiency virus-seropositive drug users. Am J Respir Crit Care Med 150:1493-1498, 1994.
  17. French N et al. 23-valent pneumococcal polysaccharide vaccine in HIV-1-infected Ugandan adults: double-blind, randomised and placebo controlled trial. The Lancet 355:2106-2111, 2000.
  18. Watera C et al. 23-Valent pneumococcal polysaccharide vaccine in HIV-infected Ugandan adults: 6-year follow-up of a clinical trial cohort. AIDS18:1210-1213, 2004.
  19. Falco V et al. Serological response to pneumococcal vaccination in HAART-treated HIV infected patients: one year follow-up study. Vaccine 24:2567-2574, 2006.
  20. Penaranda M et al. Effectiveness of polysaccharide pneumococcal vaccine in HIV-infected patients: a case-control study. Clin Infect Dis 45:e82-e87, 2007.
  21. Yamanaka H et al. Efficacy and immunologic response to influenza vaccine in HIV-1-infected patients. J Acquir Immune Defic Syndr 39: 167 - 173, 2005.
  22. Manno D et al. Risk factors and clinical characteristics associated with hospitalization for community-acquired bacterial pneumonia in HIV-positive patients according to the presence of liver cirrhosis. Infection 37:334-339, 2009.
  23. Madeddu G et al. Bacterial community-acquired pneumonia in HIV-infected patients. Curr Opin Pulm Med 16:201-207, 2010.
  24. Ibid.
  25. Centers for Disease Control and Prevention. Guidelines for prevention and treatment of opportunistic infections in HIV-infected adults and adolescents. MMWR 58:6-9, 2009.