HIV predominantly infects cells that express the CD4 surface molecule, such as T lymphocytes, monocytes and macrophages . These cells sustain the production of virions in peripheral blood and an inverse correlation between the concentration of HIV RNA in plasma and the CD4 T lymphocyte cell count has been well established . HIV infection is characterized by a progressive loss of CD4 cells, resulting in immunosuppression, which predisposes the individual to the development of opportunistic infections.
Although individual CD4 cell counts are affected by biological and laboratory variation, in the United Kingdom guidelines have been implemented to minimize these influences . The rate at which CD4 cell counts decline is one of the markers of disease progression of HIV infection, and as a result, this parameter is widely used as a guide for initiating HIV therapy and prophylaxis for opportunistic infections, and for assessing the efficacy of drug regimens in clinical trials [3–6].
The collection of CD4 cell counts at the population level supplements existing HIV surveillance systems by monitoring for trends in the immunosuppression associated with diagnosed HIV infection [7–9]. This is of increasing importance with the decline in AIDS incidence resulting from newer antiretroviral therapies. For England and Wales, CD4 cell monitoring of HIV-infected persons has been conducted through the National CD4 Surveillance Scheme, which started collecting data from a small number of centres in 1993 and has collected data nationally since 1995. The proposal for a national, laboratory based scheme was developed by a working group that included immunologists, haematologists and representatives from the Communicable Disease Surveillance Centre (CDSC), the Department of Health and the Medical Research Council. The CD4 surveillance scheme, coordinated at CDSC, has received CD4 cell count data from 55 of 71 laboratories measuring CD4 cell counts. These 55 laboratories are currently responsible for approximately 88% of CD4 cell counts performed for HIV-infected persons in England and Wales. The CD4 cell database contains records for over 35 000 individuals with approximately 247 000 CD4 cell counts, a mean of seven counts per person (range 1–99, SD 8.9).
This paper describes both the methodology of collecting the CD4 cell count data and the characteristics of adults diagnosed with HIV infection between 1990 and 1998, for whom CD4 cell counts, at or around the time of HIV diagnosis, were available through the data collection. The objectives were:
- To describe the distribution of immunosuppression in newly diagnosed adults and to determine whether the distribution has changed over time overall and within subgroups. Of particular interest were individuals who were severely immunosuppressed at diagnosis (CD4 < 200 cells per mm3), a level of immunosuppression that is an AIDS-defining condition in the USA  but not in the UK and the rest of Europe; and the proportion whose count at diagnosis fell below 350 cells/mm3, the level of immunosuppression regarded by the British HIV Association to be the level at which antiretroviral therapy should already have begun .
- To analyse the relationship between median initial CD4 cell count and individual variables such as age, sex, and HIV risk group.
- To describe the change in median CD4 cell count over time for each annual cohort of newly diagnosed individuals to determine whether there was any evidence of a treatment effect in the years after diagnosis since the introduction of combination therapy in 1996 .
All immunology and haematology laboratories known to conduct CD4 cell monitoring in England and Wales were contacted by the UK National External Quality Assessment Scheme (UK NEQAS) for Leucocyte Immunophenotyping and were invited to collaborate in the CD4 surveillance scheme. Data were collected directly from the laboratories that agreed to participate, rather than from clinical centres, to maximize the amount of data obtained and to minimize the risk of transcription errors.
Laboratories were asked to send only CD4 cell counts of those aged 15 years or older, because the distribution of CD4 cell counts is different in children . All data sent on disc were password protected and the individual surnames were replaced by ‘soundex’ codes , which were used as an additional safeguard to confidentiality. The soundex code consists of the first letter of the surname followed by three digits. Although the code is not unique to a single surname, its use in combination with the date of birth, sex and first name initial can identify multiple reports of the same individual. Because the soundex code was not always available, clinic numbers were also collected. The date of the test, the absolute CD4 cell count, CD4 cell percentage, the name of the clinician who requested the test and the hospital where the specimen was taken were also collected. Each reporting laboratory was asked to supply data going back to the start of laboratory records, followed by quarterly updates.
Identifiers for individuals on the CD4 cell database were used to search for a matching person on the UK database of diagnosed HIV infections and AIDS cases. Matching of records between the two databases also provided additional information for each person, when available, such as the date of the first positive HIV test, date of AIDS diagnosis, date of death, and probable HIV risk group.
Adults in England and Wales, diagnosed with HIV infection between 1990 and 1998, were identified from the UK database of diagnosed HIV infection and AIDS cases. Data on the HIV database are collected through voluntary confidential reporting of HIV diagnoses (by microbiologists) and of AIDS cases (by clinicians). The cohorts of newly diagnosed HIV-infected individuals were matched to an archive of the CD4 cell database created at the end of June 1999. A match confirmed that the individual recorded on the CD4 cell database was HIV infected. Matched individuals were included in the analysis if the date of their initial or first CD4 cell count was within 6 months of the HIV diagnosis date. This subgroup, defined hereafter as the study group, was then compared for age, HIV risk group, and sex to newly diagnosed adults for whom either no CD4 cell count had been identified or for whom the initial CD4 cell count was not within 6 months of diagnosis.
Descriptive statistics were used to characterize the study group. Individuals were grouped by CD4 cell count level at time of HIV diagnosis to provide an indication of the time delay between HIV infection and diagnosis. Chi-square tests were used to examine the association between the initial CD4 cell count level and calendar year periods within each risk group. Single variable quantile regression  was employed to evaluate temporal trends in the median initial CD4 cell count for the study group as a whole and the median initial CD4 cell count and median age within HIV risk groups. Departures from linearity were tested by fitting a quadratic term for time. Multi-variable quantile regression was used to determine the relationship between the median initial CD4 cell count and individual variables (age, sex, and risk category). Confidence intervals (CI) for the quantile regression estimates were obtained using 500 bootstrap replications .
The analyses were repeated including laboratory as a variable in the regression models in order to take into account laboratories entering the scheme at different times. As the addition of this variable produced the same results, for ease of presentation, all analyses are reported excluding this factor.
In a separate analysis, each cohort of newly diagnosed individuals was followed over time to examine trends in their median CD4 cell counts. The median CD4 cell count for each individual in each calendar year was calculated. Subsequently, the overall median CD4 cell count by calendar year was calculated for each cohort.
We identified 9553 adults diagnosed with HIV infection between 1990 and 1998 who had a record in the CD4 cell database of an initial CD4 cell count measured within 6 months of HIV diagnosis (Table 1). The proportion of newly diagnosed persons having CD4 cell counts recorded in the CD4 cell database within 6 months of HIV diagnosis increased from 20% in 1990 to 49% in 1998. The average number of CD4 cell counts recorded per person in the year of HIV diagnosis remained relatively stable (range 1.7 to 2.6, Table 1). The median initial CD4 cell count appeared to decrease slightly over time. This was confirmed by quantile regression analysis, which indicated that the median initial CD4 cell count at the time of HIV diagnosis decreased between 1990 and 1998 (regression coefficient, −4.8; 95% CI −7.7 to −1.9;P < 0.001). No evidence of curvature was obtained when a quadratic term for calendar time was added to the model (P = 0.40). However, despite the gradual decrease in the median initial CD4 cell count, the percentages of persons diagnosed with cell counts below 200 CD4 cells/mm3, indicating advanced disease, or above 700 CD4 cells/mm3, generally indicating more recent acquisition of HIV infection, have remained relatively stable over time (Table 1).
The study group was compared with newly diagnosed adults for whom either no CD4 cell count had been reported or for whom the initial CD4 cell count was not within 6 months of diagnosis. The mean age was 34.4 years (SD 9.1) in the 9553 persons in the study group compared with 34.2 years (SD 9.4) in the remaining 11 844 newly diagnosed adults. Women accounted for 1701 (18%) of the study group compared with 2515 (21%) in the remainder. There was a higher proportion of men who have sex with men (MSM) in the study group (65 versus 56%) and a lower proportion of individuals infected with HIV through heterosexual contact (28 versus 34%).
In order to examine the initial CD4 cell count at the time of HIV diagnosis more closely, the proportion of individuals at each CD4 cell count level by HIV risk group was also analysed by 3 year time periods (Table 2). There was no difference in the distribution of CD4 cell counts over the three time periods in MSM [χ2(6) = 7.61, P = 0.27]. This was confirmed in the quantile regression model; no association was found between the median initial CD4 cell count and the year of diagnosis (regression coefficient 1.5, 95% CI −2.1 to 5.1, P = 0.41, Fig. 1). In addition, Fig. 1 also shows that the median age at HIV diagnosis remained relatively constant with no significant change during the time period in the MSM risk group (regression coefficient 0.17, 95% CI −0.05 to 0.38, P = 0.13).
Although there was considerable heterogeneity between the three time periods in the injecting drug users (IDU) risk group [χ2(6) = 16.20, P = 0.01], there was no consistent pattern (Table 2). Figure 1 shows a downward trend in median initial CD4 cell count over time, but because of the relatively small numbers involved, the trend was not significant (regression coefficient −9.3, 95% CI −22.7 to 4.0, P = 0.17, Fig. 1). However, the median age significantly increased over the time period (regression coefficient 0.60, 95% CI 0.37 to 0.83, P < 0.001, Fig. 1).
For the heterosexual risk group, some evidence of heterogeneity was found between the three time periods [χ2(6) = 11.42, P = 0.07, Table 2]. There appeared to have been a shift in the distribution of initial CD4 cell counts towards lower values; the ratio of counts less than 350 cells/mm3 to counts greater than 350 cells/mm3 increased from 1.6 : 1 to 2.2 : 1 over the 9 year period. This trend was consistent with results obtained using the quantile regression model, in which the median initial CD4 cell count was negatively associated with the year of diagnosis (regression coefficient −7.0, 95% CI −13.3 to −0.71, P = 0.03, Fig. 1). In addition, the median age at HIV diagnosis increased over the time period examined (regression coefficient 0.43, 95% CI 0.27 to 0.59, P < 0.001, Fig. 1). When the quantile regression model for the initial CD4 cell count was adjusted for age, the initial CD4 cell count was no longer significantly associated with the year of diagnosis (regression coefficient −3.7, 95% CI −8.3 to 1.0, P = 0.11). This indicates that the decrease in the median initial CD4 cell count over time in the heterosexual risk group was partly caused by the older age of subjects in the later years of diagnosis.
Overall, the ratio of initial CD4 cell counts below 350 cells/mm3 to those above 350 cells/mm3 was lower for MSM (52 versus 48%) and IDU (53 versus 47%) than for persons exposed to HIV through heterosexual contact (67 versus 35%), Table 2.
The results of the multi-variable quantile regression analysis, used to investigate the relationship between the median initial CD4 cell count and individual variables (age at time of HIV diagnosis, sex, and risk category), are shown in Table 3. Older age was associated with a lower median initial CD4 cell count, an average decrease of 7.3 cells/mm3 per year of age (95% CI 6.6 to 8.0, P < 0.001). Although the crude median initial CD4 cell count was higher in men than in women, adjustment for risk group and age revealed that the median initial CD4 cell count at HIV diagnosis was higher in women by an average of 29.2 cells/mm3 (95% CI 6.1–52.3, P = 0.013). The median initial CD4 cell count was 101 cells/mm3 lower in persons heterosexually infected compared with MSM (230 versus 331 cells/mm3). This difference persisted after adjustment for sex and age at the time of HIV diagnosis (P < 0.001, Table 3).
The results of calculating the median CD4 cell counts for each individual in each year cohort, and subsequently determining the overall median CD4 cell count by calendar year for each cohort, are shown in Fig. 2. A clear trend is seen in which median CD4 cell counts decrease before 1996 and increase thereafter for all but one cohort (the 1993 cohort increased from 1997 onwards).
It appears from our findings that individuals have not been diagnosed any earlier in their course of HIV disease as the epidemic has progressed. Our data revealed that although the initial median CD4 cell count for newly diagnosed individuals decreased over calendar time, the proportion of individuals at each level of CD4 cell count at the time of HIV diagnosis has remained remarkably stable from 1990 to 1998. In addition, in 1998, over one-third of newly diagnosed individuals had an initial CD4 cell count below 200 CD4 cells/mm3, indicating that many infections are still being identified at a very advanced stage of HIV disease.
We determined the proportion of newly diagnosed adults at each CD4 cell count level by different HIV risk groups. In MSM, there was a relatively stable distribution of CD4 cell counts over the three time periods examined. We also found that there was no significant change in the median age and median initial CD4 cell count between 1990 and 1998 for MSM. These findings together suggest that there have continued to be many new infections entering the prevalence pool. As new diagnoses in MSM have continued at approximately 1400 per year [16,17], the steady state of median age and initial CD4 cell count at diagnosis would be consistent with an annual incidence of approximately 1400 new infections. In London, there has been an increase in the numbers of gay men who are having unprotected anal intercourse [18,19]. These results suggest that there is an urgent need to strengthen HIV prevention efforts in MSM.
For the heterosexually infected, three important findings were observed:
- The distribution of initial CD4 cell counts changed through the study period. The proportion of newly diagnosed individuals with initial CD4 cell counts below 200 cells/mm3 increased steadily between 1990 and 1998 (Table 2).
- The median age at the time of HIV diagnosis increased significantly and the median initial CD4 cell count decreased significantly from 1990 to 1998. However, after adjusting for age using regression analysis, the median initial CD4 cell count no longer decreased significantly over time (Fig. 1).
- The overall median initial CD4 cell count in heterosexuals was, on average, 101 cells/mm3 lower than for the MSM risk group (Table 3).
The above findings suggest an ageing cohort effect for the heterosexually infected. Because CD4 cell counts decrease during the progression of untreated HIV disease, the time it takes an individual to suspect they are infected and to go for an HIV test is critical in determining their initial CD4 cell count; the longer the delay between infection and diagnosis, the lower the initial count. In heterosexuals, there is a substantial delay between HIV infection and diagnosis; many individuals remain unaware of their positive HIV status until an AIDS-defining illness occurs . The overall lower median initial CD4 cell count in heterosexuals seen in our study supports these findings. The decrease in median initial CD4 cell count over time and the rising median age at HIV diagnosis are consistent with increasingly delayed diagnosis in a group mainly infected some time ago. The alternative explanation, that the incidence of HIV has intensified in older heterosexuals, does not account for the downward trend in CD4 cell count, because there is no association between CD4 cell count and age in healthy non-HIV-infected adults . Data from HIV infection reports indicate that the majority of heterosexuals diagnosed in England and Wales are people who have been exposed in sub-Saharan Africa, and probably represent migration from, or people returning having lived and worked in, high prevalence areas of Africa in the late 1980s and early 1990s [17,22,23].
For the IDU group, a trend over time was more difficult to establish. The numbers in this risk group were small, and hence the data must be interpreted with caution. The change in median initial CD4 cell count over the calendar year was not significant. However, the average decrease of 9 cells/mm3 per year and the significant increase in median age are suggestive of some ageing cohort effect in this risk group (Fig. 1). Although some recent transmission has occurred in IDU , the majority of transmission in England and Wales is thought to have occurred in the early and mid-1980s [17,25]. Therefore, it seems probable that the trend observed in this risk group is a mixture of an ageing cohort effect and the diagnosis of some recent infections.
Across all risk groups, a large percentage of individuals were diagnosed with initial CD4 cell counts below 350 CD4 cells/mm3, the level of immunosuppression regarded by the British HIV Association at which antiretroviral therapy should already have begun . In 1996–1998, the percentage of individuals with an initial count below 350 CD4 cells/mm3 was 51, 57, and 69%, respectively, for MSM, IDU, and for those heterosexually exposed. Studies conducted in the early 1990s also found similar results, with a high percentage of individuals being diagnosed with relatively low CD4 cell counts [26–28]. In addition, after adjusting for age and risk group, we also found that women have higher counts than men at HIV diagnosis, and the initial CD4 cell count decreases as age increases. These results may be explained by studies that have shown that in the HIV-negative population, women tend to have higher CD4 cell counts than men [13,20], and that this effect persists after HIV infection . As mentioned earlier, CD4 cell counts in HIV-negative adults are not affected by age , suggesting that older men and women are generally being tested later in the course of infection. With the recent advances in drug therapy, it is now even more important that individuals be diagnosed earlier in order to monitor disease progression and to ascertain the optimum time for beginning combination therapy. The data from this study indicate that there is a large number of people who could benefit from earlier diagnosis and therapy.
As the use of newer treatments in HIV-infected individuals has been shown to prolong life in several clinical trials [5,6,30], we used national CD4 cell surveillance data to determine whether the benefits of combination therapy were apparent at the population level. From 1990 to 1995, the results show a steady decline in median CD4 cell count over time for each cohort, as would be expected. However, in 1996, the year in which more effective treatments became widely available , this trend reversed and median counts increased steadily for the remaining time examined (Fig. 2). These data suggest that the benefits of combination therapy that have been observed in clinical trials are now detectable at the population level. This finding supports that of Allardice et al., who observed an increase in median CD4 cell counts in Scotland between 1996 and 1997.
The analyses revealed several important findings. However, one point of concern should be addressed. The data indicate that the proportion of newly diagnosed individuals with initial CD4 cell counts within 6 months of HIV diagnosis reported to CDSC increased steadily through the study time period. Despite this increase in CD4 cell count reporting, a large proportion of newly diagnosed persons for whom there were no counts within 6 months of HIV diagnosis were still identified on the CD4 cell database. This could have occurred for the following three reasons: (i) the counts for newly diagnosed individuals were not reported to us; (ii) the counts were reported but not within 6 months; and (iii) because of missing identifiers, the matching counts could not be identified in the CD4 cell database. As a result of the high number of individuals lacking identifiers upon initial report, we believe that missing identifiers is the most important reason for this finding. There is a possibility that individuals who are diagnosed and attend for CD4 cell monitoring within 6 months of HIV diagnosis are different from other newly diagnosed individuals. The distribution of age, sex and risk group, however, were broadly similar for both groups, except for a slight over-representation of MSM among the study group. This finding may be a reflection of health-seeking behaviour and earlier diagnosis in this risk group.
In England and Wales, a large percentage of individuals are still being diagnosed late in their course of HIV disease. Although this finding applies to all three of the major risk groups, individuals infected heterosexually continue to be particularly vulnerable to late diagnosis. Moreover, in light of our finding that MSM are diagnosed with higher median CD4 cell counts than heterosexuals, a surprisingly large number (approximately one third) of MSM are still diagnosed late in the course of clinical progression. In an era of highly active antiretroviral therapy, prompt diagnosis and the appropriate initiation of therapy are important for prolonged life and increased symptom-free survival. As transmission is still occurring, especially in MSM, better prevention efforts are also still needed. Although advances have been made in the past two decades, there is more to be achieved in effective HIV prevention, earlier detection of disease, and the provision of appropriate treatment to accomplish optimal public health benefit.
The authors are grateful to the staff in the laboratories at collaborating hospitals for providing data for the CD4 surveillance scheme: Addenbrooke's Hospital, Birmingham Heartlands Hospital, Bradford Royal Infirmary, Broomfield Hospital, Cardiff Royal Infirmary, Chelsea and Westminster Hospital, Cheltenham General Hospital, Churchill Hospital, Derby City Hospital, Derriford Hospital, Ealing Hospital, Gloucestershire Royal Hospital, Kent and Canterbury Hospital, Kings College Hospital, Kingston General Hospital, Leicester Royal Infirmary, Lewisham Hospital, Lincoln County Hospital, Maidstone Hospital, Newham General Hospital, Norfolk and Norwich Hospital, North Hampshire Hospital, North Staffordshire Hospital, Northern General Hospital NHS Trust, Poole Hospital NHS Trust, Queen Alexandra Hospital, Queen Victoria Hospital, Royal Cornwall Hospital – Treliske, Royal Free NHS Trust Hospital, Royal Hallamshire Hospital (Sheffield), Royal Liverpool University Hospital, Royal Preston Hospital, Royal Sussex County Hospital, Royal United Hospital, Royal Victoria Infirmary, Singleton Hospital, Southampton General Hospital, Southend NHS Healthcare Trust, Southmead Hospital, Southwarwickshire Hospital, St Bartholomew and London NHS Trust, St Helier Hospital, St James’ University Hospital, St Mary's Hospital (Manchester), St Mary's Hospital NHS Trust London, The General Infirmary at Leeds, The Guest Hospital, The Hillingdon Hospital, The John Radcliffe NHS Trust, The William Harvey Hospital, Torbay Hospital, University College London Hospitals, University Hospital, University Hospital, Aintree, University Hospital of Wales, Victoria Hospital, and West Hill/Joyce Green Hospital.
The authors would also like to acknowledge the following collaborators: S. Dingley, D. Bennett (Centers for Disease Control and Prevention, USA), J. Mani-Saada, P. Rogers (statistician), P. Cooper (graphic artist) and B. Martin.
1. Schnittman SM, Lane HC, Greenhouse J, Justement JS, Baseler M, Fauci AS. Preferential infection of CD4+ memory T cells by human immunodeficiency virus type 1: evidence for a role in the selective T-cell functional defects observed in infected individuals.
Proc Natl Acad Sci USA 1990, 87: 6058 –6062.
2. Feinberg MB. Changing the natural history of HIV disease.
Lancet 1996, 348: 239 –246.
3. Barnett D, Bird G, Hodges E. et al. Guidelines for the enumeration of CD4+ T lymphocytes in immunosuppressed individuals.
Clin Lab Haematol 1997, 19: 231 –241.
4. Saag MS, Holodniy M, Kuritzkes DR. et al. HIV viral load markers in clinical practice.
Nat Med 1996, 2: 625 –629.
5. Hammer SM, Katzenstein DA, Hughes MD. et al. A trial comparing nucleoside monotherapy with combination therapy in HIV-infected adults with CD4 cell counts from 200 to 500 per cubic millimeter.
N Engl J Med 1996, 335: 1081 –1090.
6. Katzenstein DA, Hammer SM, Hughes MD. et al. The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200 to 500 CD4 cells per cubic millimeter.
N Engl J Med 1995, 335: 1091 –1098.
7. CD4 Collaborative Group. CD4 surveillance in Scotland: perspectives on severe HIV-related immunodeficiency.
AIDS 1997, 11: 1509 –1517.
8. CD4 Collaborative Group. Use of monitored CD4 cell counts: predictions of the AIDS epidemic in Scotland.
AIDS 1992, 6: 213 –222.
9. Sorvillo F, Kerndt P, Cheng K. et al. Emerging patterns of HIV transmission: the value of alternative surveillance methods.
AIDS 1995, 9: 625 –629.
10. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults.
MMWR 1992, 41 (RR-17): i –19.
11. Gazzard B, Moyle G, on behalf of the BHIVA Guidelines Writing Committee. 1998 revision to the British HIV Association guidelines for antiretoviral treatment of HIV seropositive individuals.
Lancet 1998, 352: 314 –316.
12. Carpenter CC, Fischi MA, Hammer SM. et al. Antiretroviral therapy for HIV infection in 1996: recommendations of an international panel.
JAMA 1996, 276: 146 –153.
13. Bofill M, Janossy G, Lee CA. et al. Laboratory control values for CD4 and CD8 T lymphocytes.
:Implications for HIV-1 diagnosis.
Clin Exp Immunol 1992, 88: 243 –252.
14. Mortimer JY, Salathiel JA. `Soundex’ codes of surnames provide confidentiality and accuracy in national HIV database.
Commun Dis Rep Rev 1995, 5: R183 –R186.
15. Stata Corporation. Stata reference manual release 6.
College Station, Texas: Stata Press; 1999.
16. Communicable Disease Surveillance
Centre. AIDS and HIV infection in the UK: monthly report.
Commun Dis Rep Weekly 1999, 9: 121 –122.
17. Communicable Disease Surveillance
Centre. AIDS and HIV infection in the UK: monthly report.
Commun Dis Rep Weekly 1998, 8: 315 –316.
18. Nardone A, Dodds JP, Mercey DE, Johnson AM. Active surveillance of sexual behaviour among homosexual men in London.
Commun Dis Public Health 1998, 1: 197 –201.
19. Dodds J, Mercey D, Nadone A. Monitoring high-risk sexual behaviour amongst gay men in London – 1998.
Report from Department of Sexually Transmitted Diseases, Royal Free and University College London Medical School, Mortimer Market Centre (London); 1999.
20. Porter K, Wall PG, Evans BG. Factors associated with lack of awareness of HIV infection before diagnosis of AIDS.
BMJ 1993, 307: 20 –23.
21. Maini MK, Gilson RJC, Chavda N. et al. Reference ranges and sources of variability of CD4 counts in HIV-seronegative women and men.
Genitourin Med 1996, 72: 27 –31.
22. Vickers L. Trends in migration in the UK.
Population and Vital Statistics, Office for National Statistics, London 1998.
23. Haskey J. Population review: the ethnic minority and overseas born populations of Great Britain.
Population and Vital Statisitics, Office for National Statisitics, London 1997.
24. Communicable Disease Surveillance
Centre. AIDS and HIV infection in the UK: monthly report.
Commun Dis Rep Weekly 1998, 8: 357 –358.
25. Day NE, and Working Group. The incidence and prevalence of AIDS and other severe HIV disease in England and Wales for 1995 to 1999: projections using data to the end of 1994.
Commun Dis Rep Rev 1996, 6: R1 –R24.
26. Luby S, Jones J, Horan J. Using CD4 cell counts to evaluate the stages and epidemiology of HIV infection in South Carolina public clinic patients.
Am J Public Health 1994, 84: 377 –381.
27. Hutchinson CM, Wilson C, Reichart CA, Marsiglia VC, Zenilman JM, Hook EW. CD4 lymphocyte concentrations in patients with newly identified HIV infection attending STD clinics.
JAMA 1991, 266: 253 –256.
28. Brundage JF, McNeil JG, Miller RN. et al. The current distribution of CD4+ T-lymphocyte counts among adults in the United States with human immunodeficiency virus infections: estimates based on the US army.
J Acquired Immune Defic Syndr 1990, 3: 92 –94.
29. Delmas M-C, Jadand C, De Vincenzi I. et al. Gender differences in CD4+ cell counts persist after HIV-1 infection.
AIDS 1997, 11: 1071 –1073.
30. Staszewski S, Hill AM, Barlett J. et al. Reductions in HIV-1 disease progression for zidovudine/lamivudine relative to control treatments: a meta-analysis of controlled trials.
AIDS 1997, 11: 477 –483.
31. Allardice GM, McMenamin JJ, Parpia T, Gibbs J, McSharry C, Whitelaw J. The recent impact of antiretroviral combination therapy on CD4 cell counts, AIDS and death in HIV-infected persons: routine HIV surveillance in Scotland.
Int J STD AIDS 1998, 9: 561 –566.