1995
DOI: 10.1155/1996/748584
|View full text |Cite
|
Sign up to set email alerts
|

HIV‐Related Non‐Hodgkin’s Lymphoma in Calgary

Abstract: OBJECTIVE: To determine the incidence of human immunodeficiency virus (HIV) associated non-Hodgkin’s lymphoma (NHL) in a cohort of patients from a distinct geographic region (southern Alberta). The type and location of NHL as well as how it affected the survival of these patients was examined.PATIENTS AND METHODS: The Southern Alberta HIV Clinic in Calgary serves all of southern Alberta, which has an estimated population of one million. The clinic has provided primary care for 1086 patients from January 1983 t… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
1
0

Year Published

2023
2023
2023
2023

Publication Types

Select...
1

Relationship

0
1

Authors

Journals

citations
Cited by 1 publication
(1 citation statement)
references
References 24 publications
(74 reference statements)
0
1
0
Order By: Relevance
“…We defined two HIV drug resistance outcomes: the level of resistance to DTG and the presence of known drug resistance mutations (DRMs). The Stanford HIV Database version 9•0 and the Stanford HIVdb algorithm 27 were used to categorise drug resistance levels as susceptible (score below 10), potential low (10)(11)(12)(13)(14), low (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29), intermediate or high (>60). The same approach was used to assess resistance to all other antiretroviral drugs, whereby drug resistance to tenofovir alafenamide (not covered by the Stanford algorithm) was considered equal to tenofovir resistance.…”
Section: Hiv Drug Resistancementioning
confidence: 99%
“…We defined two HIV drug resistance outcomes: the level of resistance to DTG and the presence of known drug resistance mutations (DRMs). The Stanford HIV Database version 9•0 and the Stanford HIVdb algorithm 27 were used to categorise drug resistance levels as susceptible (score below 10), potential low (10)(11)(12)(13)(14), low (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29), intermediate or high (>60). The same approach was used to assess resistance to all other antiretroviral drugs, whereby drug resistance to tenofovir alafenamide (not covered by the Stanford algorithm) was considered equal to tenofovir resistance.…”
Section: Hiv Drug Resistancementioning
confidence: 99%