On 2016 Mar 01, Ram Dessau commented:

Residual symptoms after Lyme neuroborreliosis. The prevalence is unknown? Comment by Ram B. Dessau

Department clinical microbiology Slagelse Hospital, Region Zealand Ingemannsvej 46, DK4200, Slagelse, Denmark ramd@regionsjaelland.dk

Dersch et al. have performed a systematic review and meta-analysis on the prevalence and spectrum of residual symptoms in patients after treatment for Lyme neuroborreliosis (LNB)[1]. 5579 bibliographic records were screened and 38 studies were included. An overview of treatment studies reporting residual symptoms is given. The follow up time is about one year concerning most of the studies. The conclusion is that 28% of LNB patients may experience symptoms after treatment. This conclusion may be misinterpreted as associated to LNB or caused by LNB. The 38 treatment studies were not designed to elucidate the association of LNB with residual symptoms, as control groups without LNB were not included.

Many of the cited symptoms like sensory disturbances, cognitive disturbances, pain and headache are non-specific and even common in the population at large. Thus, it would be of paramount importance to compare the prevalence of events to a control group using the same methods. Only a higher frequency of residual symptoms in the LNB group may indicate association with LNB. Due to the design of the included studies the residual symptoms could as well be due to other causes such as adverse effects of treatment [2].

Reporting the gross proportion of events may be misleading for other reasons. The load of residual symptoms will depend on not only the number of events for each symptom, but also the number of recorded items and the length of time, where symptoms may occur. The more questions you ask and the higher the gross rate of symptoms will become and eventually the rate of symptoms may approach 100%. Especially if there is also a long time span for the observations.

Along this line of reasoning Dersch et al. also misinterpreted the studies including a control group[1]. The results of these studies do not “remain inconclusive”, but show comparable results, even if there are differences. These differences are not very large and a substantial proportion of the controls also have a similar level of problems. For example the mean physical component FS36 score in LNB was 44(9) compared to 51(6) in controls[3]. A significant difference in mean score was found, but the size of the standard deviation show that the distribution of scores in the two groups have a large overlap. Given that the 95% interval is about 2 SD then FS36 score in LNB is 26-62 and in controls 39-63. Even if LNB patients have “significantly” lower FS36 score than controls, this is not a large difference. Thus a weak or absent association alone could explain that some studies do not report statistical differences and some do due to random variation. Thus, residual symptoms associated with LNB are probably quite rare, if at all.

Grouping together any symptoms without considering a variable relevance to LNB is problematic. For example, it was found in Swedish children that residual symptoms were persistent nerve defects such as facial palsy, but not nonspecific subjective symptoms [4]. Thus, it would be important to distinguish different types of symptoms and their severity. The study by Dersch et al. adds to a large volume of uncontrolled case reports and case series overstating the possible risk of symptoms associated with Lyme borreliosis[5]. This may contribute to unnecessary anxiety about Lyme borreliosis.

It is acknowledged that Dersch et al [1] do state important reservations about especially the "possible" case definitions, as patients with other conditions could contribute to the high load of reported symptomps.

Conclusion The prevalence of residual symptoms associated with LNB after treatment may not be concluded from the study by Dersch et al [1] because control groups are missing. A meaningfull interpretation is not possible. The authors should have focused on the available controlled studies instead.

Conflict of interests: None to declare.

References

1.Dersch R, Sommer H, Rauer S, Meerpohl JJ. Prevalence and spectrum of residual symptoms in Lyme neuroborreliosis after pharmacological treatment: a systematic review. J Neurol 2015 Oct 12.

2.Dersch R, Freitag MH, Schmidt S, Sommer H, Rauer S, Meerpohl JJ. Efficacy and safety of pharmacological treatments for acute Lyme neuroborreliosis - a systematic review. Eur J Neurol 2015 Sep;22:1249-59.

3.Eikeland R, Mygland A, Herlofson K, Ljostad U. European neuroborreliosis: quality of life 30 months after treatment. Acta Neurol Scand 2011 Nov;124:349-54.

4.Skogman BH, Glimaker K, Nordwall M, Vrethem M, Odkvist L, Forsberg P. Long-term clinical outcome after Lyme neuroborreliosis in childhood. Pediatrics 2012 Aug;130:262-9.

5.Baker CJ, et al. Final report of the Lyme disease review panel of the infectious diseases society of America (http://www.idsociety.org).2010.

On 2016 Mar 01, Ram Dessau commented:

Residual symptoms after Lyme neuroborreliosis. The prevalence is unknown? Comment by Ram B. Dessau

Department clinical microbiology Slagelse Hospital, Region Zealand Ingemannsvej 46, DK4200, Slagelse, Denmark ramd@regionsjaelland.dk

Dersch et al. have performed a systematic review and meta-analysis on the prevalence and spectrum of residual symptoms in patients after treatment for Lyme neuroborreliosis (LNB)[1]. 5579 bibliographic records were screened and 38 studies were included. An overview of treatment studies reporting residual symptoms is given. The follow up time is about one year concerning most of the studies. The conclusion is that 28% of LNB patients may experience symptoms after treatment. This conclusion may be misinterpreted as associated to LNB or caused by LNB. The 38 treatment studies were not designed to elucidate the association of LNB with residual symptoms, as control groups without LNB were not included.

Many of the cited symptoms like sensory disturbances, cognitive disturbances, pain and headache are non-specific and even common in the population at large. Thus, it would be of paramount importance to compare the prevalence of events to a control group using the same methods. Only a higher frequency of residual symptoms in the LNB group may indicate association with LNB. Due to the design of the included studies the residual symptoms could as well be due to other causes such as adverse effects of treatment [2].

Reporting the gross proportion of events may be misleading for other reasons. The load of residual symptoms will depend on not only the number of events for each symptom, but also the number of recorded items and the length of time, where symptoms may occur. The more questions you ask and the higher the gross rate of symptoms will become and eventually the rate of symptoms may approach 100%. Especially if there is also a long time span for the observations.

Along this line of reasoning Dersch et al. also misinterpreted the studies including a control group[1]. The results of these studies do not “remain inconclusive”, but show comparable results, even if there are differences. These differences are not very large and a substantial proportion of the controls also have a similar level of problems. For example the mean physical component FS36 score in LNB was 44(9) compared to 51(6) in controls[3]. A significant difference in mean score was found, but the size of the standard deviation show that the distribution of scores in the two groups have a large overlap. Given that the 95% interval is about 2 SD then FS36 score in LNB is 26-62 and in controls 39-63. Even if LNB patients have “significantly” lower FS36 score than controls, this is not a large difference. Thus a weak or absent association alone could explain that some studies do not report statistical differences and some do due to random variation. Thus, residual symptoms associated with LNB are probably quite rare, if at all.

Grouping together any symptoms without considering a variable relevance to LNB is problematic. For example, it was found in Swedish children that residual symptoms were persistent nerve defects such as facial palsy, but not nonspecific subjective symptoms [4]. Thus, it would be important to distinguish different types of symptoms and their severity. The study by Dersch et al. adds to a large volume of uncontrolled case reports and case series overstating the possible risk of symptoms associated with Lyme borreliosis[5]. This may contribute to unnecessary anxiety about Lyme borreliosis.

It is acknowledged that Dersch et al [1] do state important reservations about especially the "possible" case definitions, as patients with other conditions could contribute to the high load of reported symptomps.

Conclusion The prevalence of residual symptoms associated with LNB after treatment may not be concluded from the study by Dersch et al [1] because control groups are missing. A meaningfull interpretation is not possible. The authors should have focused on the available controlled studies instead.

Conflict of interests: None to declare.

References

1.Dersch R, Sommer H, Rauer S, Meerpohl JJ. Prevalence and spectrum of residual symptoms in Lyme neuroborreliosis after pharmacological treatment: a systematic review. J Neurol 2015 Oct 12.

2.Dersch R, Freitag MH, Schmidt S, Sommer H, Rauer S, Meerpohl JJ. Efficacy and safety of pharmacological treatments for acute Lyme neuroborreliosis - a systematic review. Eur J Neurol 2015 Sep;22:1249-59.

3.Eikeland R, Mygland A, Herlofson K, Ljostad U. European neuroborreliosis: quality of life 30 months after treatment. Acta Neurol Scand 2011 Nov;124:349-54.

4.Skogman BH, Glimaker K, Nordwall M, Vrethem M, Odkvist L, Forsberg P. Long-term clinical outcome after Lyme neuroborreliosis in childhood. Pediatrics 2012 Aug;130:262-9.

5.Baker CJ, et al. Final report of the Lyme disease review panel of the infectious diseases society of America (http://www.idsociety.org).2010.